Parameter Types

IParameter

class IParameter

Subclassed by visionflow::param::BinaryPacks, visionflow::param::BinaryTrainingParameters, visionflow::param::IAveragePrecisionList, visionflow::param::IPrecisionEvaluation, visionflow::param::ITableList, visionflow::param::LabelClasses, visionflow::param::SchemableParameter, visionflow::param::ViewTransformParameterList

ISchemable

struct ISchemable

Json Schemable Class API.

A Json Schemable Class is a class that can be serialized to/deserialized from Json and can be validated by Json Schema.

Subclassed by visionflow::param::AssemblyVerificationLayOutArea, visionflow::param::AssemblyVerificationMaxInputSize, visionflow::param::AssemblyVerificationModelParameters, visionflow::param::AssemblyVerificationTargetParameters, visionflow::param::AxialSideLengthRange, visionflow::param::ClassificationInputShape, visionflow::param::DataAugmentation, visionflow::param::DetectionInputShape, visionflow::param::DetectionTrainingStrategy, visionflow::param::ExampleAugRotate, visionflow::param::ExampleAugShift, visionflow::param::FilterScript, visionflow::param::GeoAugDistortion, visionflow::param::GeoAugScale, visionflow::param::GeoAugShift, visionflow::param::GeoAugSlightRotation, visionflow::param::GeometryAugmentation, visionflow::param::ImageAugBrightness, visionflow::param::ImageAugColorFilter, visionflow::param::ImageAugContrast, visionflow::param::ImageAugIlluminationGradient, visionflow::param::ImageAugNoise, visionflow::param::ImageAugSmoothingOrSharpening, visionflow::param::ImageAugmentation, visionflow::param::KeyPointNode, visionflow::param::LocationMaxInputSize, visionflow::param::LocationModelParameters, visionflow::param::LocationTargetParameters, visionflow::param::LossCurve, visionflow::param::MultiNameKeyPointNode, visionflow::param::NodeMatchTemplate, visionflow::param::OCRAnnulusParameters, visionflow::param::OCRNodeTemplate, visionflow::param::OCRStringTemplate, visionflow::param::PropertyObjectId, visionflow::param::RegularExpression, visionflow::param::SchemableParameter, visionflow::param::SegmentationImageSplit, visionflow::param::SegmentationInputShape, visionflow::param::SegmentationModelParameters, visionflow::param::SegmentationTrainingSampleStrategy, visionflow::param::SideLengthRange, visionflow::param::SingleClassPolygonsFilterParameters, visionflow::param::UnsuperClassificationInputShape, visionflow::param::UnsuperDefectRadius, visionflow::param::UnsuperSegmentationInputShape, visionflow::param::UnsuperSegmentationModelParameters, visionflow::param::UnsuperSegmentationSamplingParameters

Public Functions

virtual const std::string &schema_str() const = 0

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const = 0

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

const void* The nlohmann::json object pointer

virtual visionflow::Buffer to_json() const = 0

Serialize the parameter group to schema-validatable json string.

Returns:

visionflow::Buffer the schema-validatable json string.

virtual ISchemable &from_json(const visionflow::Buffer &json_str) = 0

Deserialize the parameter group from schema-validatable json string.

See also

validate(const std::string &json_str, std::string *error_msg = nullptr)

Parameters:

json_str – [in] the schema-validatable json string.

Throws:

• visionflow::excepts::InvalidJson – if the json string is invalid.

• visionflow::excepts::JsonSchemaValidationError – if the json string is not serialized from this parameter type or validation the json with json schema failed.

Returns:

ISchemable& the reference of the parameter.

bool validate(const visionflow::Buffer &json_str, std::string *error_msg = nullptr) const

Validate the parameter json string with json schema.

Parameters:

• json_str – [in] the json string.

• error_msg – [out] the error message if the json string is invalid, otherwise empty.

Throws:

visionflow::excepts::InvalidJson – if the json string is invalid.

Returns:

bool true if the json string is valid, otherwise false.

SchemableParameter

struct SchemableParameter : public virtual visionflow::param::IParameter, public virtual visionflow::param::ISchemable

SchemableParameter feature interface for IParameter classes.

Subclassed by visionflow::param::AssemblyVerificationFilterParameters, visionflow::param::AssemblyVerificationTemplates, visionflow::param::AssemblyVerificationTrainingParameters, visionflow::param::BaseColor, visionflow::param::ClassificationInferParameters, visionflow::param::ClassificationTrainingParameters, visionflow::param::DetectionInferParameters, visionflow::param::DetectionTrainingParameters, visionflow::param::ExampleAugments, visionflow::param::FeatureMapFilterParameters, visionflow::param::ImageMean, visionflow::param::InferenceBatchSize, visionflow::param::InputImageParam, visionflow::param::IntegrationClassifyParameter, visionflow::param::LocationFilterParameters, visionflow::param::LocationTemplates, visionflow::param::LocationTrainingParameters, visionflow::param::OCRFilterParameters, visionflow::param::OCRInferParameters, visionflow::param::OCRTemplates, visionflow::param::OCRTrainingParameters, visionflow::param::OCRUniversalModelParameters, visionflow::param::PolygonsFilterParameters, visionflow::param::PropertyObjectIdSet, visionflow::param::SampleRecommendationParameter, visionflow::param::SegmentationTrainingParameters, visionflow::param::TRTCalibParameters, visionflow::param::TrainingLog, visionflow::param::TrainingSetRecommendParameter, visionflow::param::UnsuperClassificationInferenceParameters, visionflow::param::UnsuperClassificationTrainingParameters, visionflow::param::UnsuperSegmentationInferenceParameters, visionflow::param::UnsuperSegmentationTrainingParameters, visionflow::param::ViewFilterParameters

Customize Parameters

class BinaryPacks : public visionflow::param::IParameter

Parameter to store list of binary data with name.

Public Functions

void insert(const std::string &key, const visionflow::Buffer &buffer)

Insert a new binary data into the BinaryPacks. Old data with the same key will be replaced.

See also

visionflow::Buffer

See also

get(const std::string &key)

Note

The data in the buffer will not be copied, Which means that if you modify the data content in the Buffer later, the data in the inserted Buffer will also be modified. If this is not the behavior you expect, you can call visionflow::Buffer::clone_data() explicitly like below:

binary_packs.insert("my-data", buffer.clone_data());

Parameters:

• key – The key of the binary data.

• buffer – The binary data.

void remove(const std::string &key)

Remove binary data specified by the key. Do nothing if the key not found.

Parameters:

key – The key of the binary data to be removed.

bool contains(const std::string &key) const

Check if the binary data exists in the container.

Parameters:

key – The key of the binary data

Returns:

True if exists. Else false.

const visionflow::Buffer &get(const std::string &key) const

Get the binary data specified by the key.

See also

insert(const std::string &key, const visionflow::Buffer &buffer) for more details.

Note

Do not modify the data in the buffer with out clone,

Parameters:

key –

Returns:

const visionflow::Buffer & The reference of the binary data.

std::vector<std::string> keys() const

Get all keys in the BinaryPacks.

Returns:

std::vector<std::string> The keys list.

void clear()

Clear the container.

InputImage Parameters

enum visionflow::param::ColorType

Values:

enumerator kGray = 1

enumerator kBGR = 3

class BaseColor : public visionflow::param::SchemableParameter

BaseColor Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const ColorType &get_color() const

Get base_color.

Convert all images into Gray or BGR color space

See also

set_color()

Returns:

const ColorType & base_color

BaseColor &set_color(const ColorType &color)

Set base_color with ColorType value.

Convert all images into Gray or BGR color space

See also

get_color()

Parameters:

color – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

BaseColor& the reference of this object.

class ImageMean : public visionflow::param::SchemableParameter

ImageMean Parameter class generated by jinja2 automatically.

Image mean parameters

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const std::vector<double> &get_values() const

Get image mean.

image mean values ordered same as image channels

See also

set_values()

Returns:

const std::vector<double> & image mean

ImageMean &set_values(const std::vector<double> &values)

Set image mean with std::vector<double> value.

image mean values ordered same as image channels

See also

get_values()

Parameters:

values – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageMean& the reference of this object.

double get_values(size_t index) const

Get value in image mean with index.

Warning

The index must be less than get_values_size(). otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

double value in image mean at index.

size_t get_values_size() const

Get the size of image mean.

Returns:

size_t the size of image mean

class InputImageParam : public visionflow::param::SchemableParameter

InputImageParam Parameter class generated by jinja2 automatically.

输入图像相关参数，用于控制工程的输入图像格式规范.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int get_visual_size() const

Get 可视图象数.

一个visionflow.Image允许包含的可视图象数量, 多数情况下, 使用默认值1即可。若需要将多张图像当作一张图像处理, 可以设置为大于1的值。

See also

set_visual_size()

Returns:

int 可视图象数

InputImageParam &set_visual_size(int visual_size)

Set 可视图象数 with int value.

一个visionflow.Image允许包含的可视图象数量, 多数情况下, 使用默认值1即可。若需要将多张图像当作一张图像处理, 可以设置为大于1的值。

See also

get_visual_size()

Parameters:

visual_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

InputImageParam& the reference of this object.

int get_thumbnail_image_size() const

Get 缩略图尺寸大小.

缩略图长边的最大尺寸，最长边默认为512，最大不能超过512.

See also

set_thumbnail_image_size()

Returns:

int 缩略图尺寸大小

InputImageParam &set_thumbnail_image_size(int thumbnail_image_size)

Set 缩略图尺寸大小 with int value.

缩略图长边的最大尺寸，最长边默认为512，最大不能超过512.

See also

get_thumbnail_image_size()

Parameters:

thumbnail_image_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

InputImageParam& the reference of this object.

class LabelClasses : public visionflow::param::IParameter

LabelClasses is a parameter to manage label classes.

Public Functions

void add(const std::string &label_name)

Add a label name into the label classes container.

Parameters:

label_name – The label name to be added.

void remove(const std::string &label_name)

Remove a label name from the label classes container.

Parameters:

label_name – The label name to be removed.

size_t size() const

Get the number of label classes.

Returns:

The number of label classes.

inline bool empty() const

Check if the container is empty.

Returns:

bool True if the container is empty.

bool contains(const std::string &label_name) const

Check if the label classes container contains the label name.

Parameters:

label_name – The label name to be checked.

Returns:

True if the label classes container contains the label name.

std::vector<std::string> get() const

Get all label names in the label classes container as a vector.

Returns:

std::vector<std::string> The label names in the label classes

void clear()

Clear the label classes container.

EvaluationMetrics Parameters

enum visionflow::param::MetricsEvaluationType

Values:

enumerator kImage = 0

evaluation metrics of all images.

enumerator kAllView = 1

evaluation metrics of all views.

enumerator kTestView = 2

evaluation metrics of all test set views.

enumerator kTrainView = 3

evaluation metrics of all train set views.

enumerator kAllRegion = 4

evaluation metrics of all regions.

enumerator kTestRegion = 5

evaluation metrics of all test set regions.

enumerator kTrainRegion = 6

evaluation metrics of all train set regions.

enumerator kAllPixel = 7

evaluation metrics of all pixels.

enumerator kTestPixel = 8

evaluation metrics of all test set pixels.

enumerator kTrainPixel = 9

evaluation metrics of all train set pixels.

enum visionflow::param::PrecisionEvaluationType

Values:

enumerator kAngle = 0

Precision evaluation of angle.

enumerator kPosition = 1

Precision evaluationfor of position.

class Table

Public Functions

size_t rows() const

The number of rows.

Returns:

size_t

size_t cols() const

The number of columns.

Returns:

size_t

const std::vector<std::string> &row_header() const

Get the row header.

Returns:

std::vector<std::string> The row header.

const std::vector<std::string> &col_header() const

Get the column header.

Returns:

std::vector<std::string> The column header.

std::vector<int> row_data(size_t row) const

Get the data of one row.

Parameters:

row – The index of rows.

Throws:

excepts::DataNotFound – If no such the row.

Returns:

std::vector<int> The data of one row.

std::vector<int> col_data(size_t col) const

Get the data of one column.

Parameters:

col – The index of columns.

Throws:

excepts::DataNotFound – If no such the column.

Returns:

std::vector<int> The data of one column.

int data(size_t row, size_t col) const

Get the data of the specific position in the table.

Parameters:

• row – The index of rows.

• col – The index of columns.

Throws:

excepts::DataNotFound – If no such the row or column.

Returns:

int

int &data(size_t row, size_t col)

Get the reference of the data of the specific position in the table.

Parameters:

• row – The index of rows.

• col – The index of columns.

Throws:

excepts::DataNotFound – If no such the row or column.

Returns:

NO_DISCARD&

Table &reset_header(const std::vector<std::string> &row_header, const std::vector<std::string> &col_header)

Reset the header of the table.

Note

If no header is set at construction time，then the first task is to reset header. The number of rows and columns of this matrix will be initialized by the length of the row and column header. If the header has been set before, the reset header will clear the previous data, please be careful.

Parameters:

• row_header – The header of row.

• col_header – The header of column.

Returns:

Table&

Table &set_row_data(size_t row, const std::vector<int> &data)

Set the data of one row.

Parameters:

• row – The index of rows.

• data –

Throws:

• excepts::DataNotFound – If no such the row.

• excepts::InvalidArgument – The data size does not meet the row number requirement of the table.

Returns:

Table&

Table &set_col_data(size_t col, const std::vector<int> &data)

Set the data of one column.

Parameters:

• col – The index of columns.

• data –

Throws:

• excepts::DataNotFound – If no such the column.

• excepts::InvalidArgument – The data size does not meet the column number requirement of the table.

Returns:

Table&

Table &set_data(size_t row, size_t col, int data)

Set the data of the specific position in the table.

Parameters:

• row – The index of rows.

• col – The index of columns.

• data –

Throws:

excepts::DataNotFound – If no such the row or column.

Returns:

Table&

class DataMetrics

Public Functions

float maximum() const

Get the maximum.

Returns:

NO_DISCARD

float minimum() const

Get the minimum.

Returns:

NO_DISCARD

float average() const

Get the average.

Returns:

NO_DISCARD

float standard_deviation() const

Get the standard deviation.

Returns:

NO_DISCARD

DataMetrics &set_maximum(float value)

Set the maximum.

Parameters:

value –

Returns:

DataMetrics&

DataMetrics &set_minimum(float value)

Set the minimum.

Parameters:

value –

Returns:

DataMetrics&

DataMetrics &set_average(float value)

Set the average.

Parameters:

value –

Returns:

DataMetrics&

DataMetrics &set_standard_deviation(float value)

Set the standard deviation.

Parameters:

value –

Returns:

DataMetrics&

class AveragePrecision

Public Functions

AveragePrecision &set_average_precision(const std::string &key, float average_precision)

Set the specified average precision.

Parameters:

• key –

• average_precision –

float average_precision(const std::string &key) const

Get the specified average precision.

Parameters:

key –

Returns:

NO_DISCARD

float mean_average_precision() const

Get the mean average precision.

Returns:

NO_DISCARD

std::vector<std::string> keys() const

Get the keys.

Returns:

NO_DISCARD const&

float precision() const

Get the precision.

Note

Returns -1, if there is no precision property.

Returns:

NO_DISCARD

float recall() const

Get the recall.

Note

Returns -1, if there is no recall property.

Returns:

NO_DISCARD

AveragePrecision &set_precision(float precision)

Set the precision.

Parameters:

precision –

Returns:

AveragePrecision&

AveragePrecision &set_recall(float recall)

Set the recall.

Parameters:

recall –

Returns:

AveragePrecision&

struct IPrecisionEvaluation : public virtual visionflow::param::IParameter

IPrecisionEvaluation feature interface for IParameter classes.

Subclassed by visionflow::param::LocationModelEvaluationMetrics

Public Functions

virtual std::set<PrecisionEvaluationType> precision_evaluation_keys() const = 0

Get the types of precision evaluation in this map.

Returns:

std::set<PrecisionEvaluationType>

virtual const DataMetrics &precision_evaluation(PrecisionEvaluationType type) const = 0

Get the precision evaluation with the type.

Parameters:

type –

Returns:

NO_DISCARD

virtual DataMetrics &precision_evaluation(PrecisionEvaluationType type) = 0

Get the reference precision evaluation with the type.

Parameters:

type –

Returns:

NO_DISCARD

virtual IPrecisionEvaluation &set_precision_evaluation(PrecisionEvaluationType type, const DataMetrics &data_metrics) = 0

Set the precision evaluation with the type.

Parameters:

• type –

• data_metrics –

Returns:

IPrecisionEvaluation&

struct ITableList : public virtual visionflow::param::IParameter

ITableList feature interface for IParameter classes.

Subclassed by visionflow::param::LocationModelEvaluationMetrics, visionflow::param::ModelEvaluationMetrics

Public Functions

virtual std::vector<MetricsEvaluationType> table_keys() const = 0

Get the types of table in this map.

Returns:

std::vector<MetricsEvaluationType>

virtual const Table &table(MetricsEvaluationType type) const = 0

Get the table with the type.

Parameters:

type –

Returns:

NO_DISCARD

virtual Table &table(MetricsEvaluationType type) = 0

Get the reference table with the type.

Parameters:

type –

Returns:

NO_DISCARD

virtual ITableList &set_table(MetricsEvaluationType type, const Table &table) = 0

Set the table.

Parameters:

• type –

• table –

Returns:

ITableList&

struct IAveragePrecisionList : public virtual visionflow::param::IParameter

IAveragePrecisionList feature interface for IParameter classes.

Subclassed by visionflow::param::LocationModelEvaluationMetrics, visionflow::param::ModelEvaluationMetrics

Public Functions

virtual std::vector<MetricsEvaluationType> average_precision_keys() const = 0

Get the types of average_precision in this map.

Returns:

std::vector<MetricsEvaluationType>

virtual const AveragePrecision &average_precision(MetricsEvaluationType type) const = 0

Get the average precision with the type.

Parameters:

type –

Returns:

NO_DISCARD

virtual AveragePrecision &average_precision(MetricsEvaluationType type) = 0

Get the reference average precision with the type.

Parameters:

type –

Returns:

NO_DISCARD

virtual IAveragePrecisionList &set_average_precision(MetricsEvaluationType type, const AveragePrecision &average_precision) = 0

Set the average precision.

Parameters:

• type –

• average_precision –

Returns:

IAveragePrecisionList&

class ModelEvaluationMetrics : public virtual visionflow::param::ITableList, public virtual visionflow::param::IAveragePrecisionList

Public Functions

virtual std::vector<MetricsEvaluationType> table_keys() const override

Get the types of table in this map.

Returns:

std::vector<MetricsEvaluationType>

virtual const Table &table(MetricsEvaluationType type) const override

Get the table with the type.

Parameters:

type –

Throws:

excepts::DataNotFound – The table with the type does not exist.

Returns:

NO_DISCARD

virtual Table &table(MetricsEvaluationType type) override

Get the reference table with the type.

Parameters:

type –

Throws:

excepts::DataNotFound – The table with the type does not exist.

Returns:

NO_DISCARD

virtual ModelEvaluationMetrics &set_table(MetricsEvaluationType type, const Table &table) override

Set the table.

Parameters:

• type –

• table –

Returns:

ITableList&

virtual std::vector<MetricsEvaluationType> average_precision_keys() const override

Get the types of average_precision in this map.

Returns:

std::vector<MetricsEvaluationType>

virtual const AveragePrecision &average_precision(MetricsEvaluationType type) const override

Get the average precision with the type.

Parameters:

type –

Throws:

excepts::DataNotFound – The average precision with the type does not exist.

Returns:

NO_DISCARD

virtual AveragePrecision &average_precision(MetricsEvaluationType type) override

Get the reference average precision with the type.

Parameters:

type –

Throws:

excepts::DataNotFound – The average precision with the type does not exist.

Returns:

NO_DISCARD

virtual ModelEvaluationMetrics &set_average_precision(MetricsEvaluationType type, const AveragePrecision &average_precision) override

Set the average precision.

Parameters:

• type –

• average_precision –

Returns:

IAveragePrecisionList&

class LocationModelEvaluationMetrics : public virtual visionflow::param::ITableList, public virtual visionflow::param::IPrecisionEvaluation, public virtual visionflow::param::IAveragePrecisionList

Public Functions

virtual std::vector<MetricsEvaluationType> table_keys() const override

Get the types of table in this map.

Returns:

std::vector<MetricsEvaluationType>

virtual const Table &table(MetricsEvaluationType type) const override

Get the table with the type.

Parameters:

type –

Throws:

excepts::DataNotFound – The table with the type does not exist.

Returns:

NO_DISCARD

virtual Table &table(MetricsEvaluationType type) override

Get the reference table with the type.

Parameters:

type –

Throws:

excepts::DataNotFound – The table with the type does not exist.

Returns:

NO_DISCARD

virtual LocationModelEvaluationMetrics &set_table(MetricsEvaluationType type, const Table &table) override

Set the table.

Parameters:

• type –

• table –

Returns:

ITableList&

virtual std::set<PrecisionEvaluationType> precision_evaluation_keys() const override

Get the types of precision evaluation in this map.

Returns:

std::set<PrecisionEvaluationType>

virtual const DataMetrics &precision_evaluation(PrecisionEvaluationType type) const override

Get the precision evaluation with the type.

Parameters:

type –

Throws:

excepts::DataNotFound – The precision evaluation with the type does not exist.

Returns:

NO_DISCARD

virtual DataMetrics &precision_evaluation(PrecisionEvaluationType type) override

Get the reference precision evaluation with the type.

Parameters:

type –

Throws:

excepts::DataNotFound – The precision evaluation with the type does not exist.

Returns:

NO_DISCARD

virtual LocationModelEvaluationMetrics &set_precision_evaluation(PrecisionEvaluationType type, const DataMetrics &data_metrics) override

Set the precision evaluation with the type.

Parameters:

• type –

• data_metrics –

Returns:

LocationModelEvaluationMetrics&

virtual std::vector<MetricsEvaluationType> average_precision_keys() const override

Get the types of average_precision in this map.

Returns:

std::vector<MetricsEvaluationType>

virtual const AveragePrecision &average_precision(MetricsEvaluationType type) const override

Get the average precision with the type.

Parameters:

type –

Throws:

excepts::DataNotFound – The average precision with the type does not exist.

Returns:

NO_DISCARD

virtual AveragePrecision &average_precision(MetricsEvaluationType type) override

Get the reference average precision with the type.

Parameters:

type –

Throws:

excepts::DataNotFound – The average precision with the type does not exist.

Returns:

NO_DISCARD

virtual LocationModelEvaluationMetrics &set_average_precision(MetricsEvaluationType type, const AveragePrecision &average_precision) override

Set the average precision.

Parameters:

• type –

• average_precision –

Returns:

IAveragePrecisionList&

DataAugmentation Parameters

class GeoAugSlightRotation : public visionflow::param::ISchemable

GeoAugSlightRotation Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable.

Enable to augment training data with slight rotation

See also

set_enable()

Returns:

bool Enable

GeoAugSlightRotation &set_enable(bool enable)

Set Enable with bool value.

Enable to augment training data with slight rotation

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugSlightRotation& the reference of this object.

const std::vector<double> &get_range() const

Get Angle Range.

Slight rotation angle range (min, max)

See also

set_range()

Returns:

const std::vector<double> & Angle Range

GeoAugSlightRotation &set_range(const std::vector<double> &range)

Set Angle Range with std::vector<double> value.

Slight rotation angle range (min, max)

See also

get_range()

Parameters:

range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugSlightRotation& the reference of this object.

bool range_contains(double value) const

Check if the Angle Range contains the value.

Returns:

bool true if the Angle Range contains the value, otherwise false.

double get_range_left() const

Get left point value of Angle Range.

See also

set_range_right()

Returns:

const std::vector<double> & left point value of Angle Range

double get_range_right() const

Get the right point value of Angle Range.

See also

set_range_left()

Returns:

const std::vector<double> & the right point value of Angle Range

GeoAugSlightRotation &set_range_left(double range_left)

Set left point value of Angle Range with std::vector<double> value.

Slight rotation angle range (min, max)

See also

get_range_left()

Parameters:

range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugSlightRotation& the reference of this object.

GeoAugSlightRotation &set_range_right(double range_right)

Set the right point value of Angle Range with std::vector<double> value.

Slight rotation angle range (min, max)

See also

get_range_right()

Parameters:

range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugSlightRotation& the reference of this object.

int get_step() const

Get Angle Step.

Slight rotation angle step

See also

set_step()

Returns:

int Angle Step

GeoAugSlightRotation &set_step(int step)

Set Angle Step with int value.

Slight rotation angle step

See also

get_step()

Parameters:

step – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugSlightRotation& the reference of this object.

class GeoAugShift : public visionflow::param::ISchemable

GeoAugShift Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable.

Enable to randomly shift the training data horizontally and vertically by a certain proportion

See also

set_enable()

Returns:

bool Enable

GeoAugShift &set_enable(bool enable)

Set Enable with bool value.

Enable to randomly shift the training data horizontally and vertically by a certain proportion

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugShift& the reference of this object.

double get_shift_vertical() const

Get Vertical Shift.

Vertical shift

See also

set_shift_vertical()

Returns:

double Vertical Shift

GeoAugShift &set_shift_vertical(double shift_vertical)

Set Vertical Shift with double value.

Vertical shift

See also

get_shift_vertical()

Parameters:

shift_vertical – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugShift& the reference of this object.

double get_shift_horizontal() const

Get Horizontal Shift.

Horizontal shift

See also

set_shift_horizontal()

Returns:

double Horizontal Shift

GeoAugShift &set_shift_horizontal(double shift_horizontal)

Set Horizontal Shift with double value.

Horizontal shift

See also

get_shift_horizontal()

Parameters:

shift_horizontal – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugShift& the reference of this object.

class GeoAugScale : public visionflow::param::ISchemable

GeoAugScale Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable.

Enable to randomly scale the training data by a certain proportion

See also

set_enable()

Returns:

bool Enable

GeoAugScale &set_enable(bool enable)

Set Enable with bool value.

Enable to randomly scale the training data by a certain proportion

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugScale& the reference of this object.

const std::vector<double> &get_range() const

Get Scale Range.

Scale range (min, max)

See also

set_range()

Returns:

const std::vector<double> & Scale Range

GeoAugScale &set_range(const std::vector<double> &range)

Set Scale Range with std::vector<double> value.

Scale range (min, max)

See also

get_range()

Parameters:

range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugScale& the reference of this object.

bool range_contains(double value) const

Check if the Scale Range contains the value.

Returns:

bool true if the Scale Range contains the value, otherwise false.

double get_range_left() const

Get left point value of Scale Range.

See also

set_range_right()

Returns:

const std::vector<double> & left point value of Scale Range

double get_range_right() const

Get the right point value of Scale Range.

See also

set_range_left()

Returns:

const std::vector<double> & the right point value of Scale Range

GeoAugScale &set_range_left(double range_left)

Set left point value of Scale Range with std::vector<double> value.

Scale range (min, max)

See also

get_range_left()

Parameters:

range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugScale& the reference of this object.

GeoAugScale &set_range_right(double range_right)

Set the right point value of Scale Range with std::vector<double> value.

Scale range (min, max)

See also

get_range_right()

Parameters:

range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugScale& the reference of this object.

class GeoAugDistortion : public visionflow::param::ISchemable

GeoAugDistortion Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable.

Enable to randomly distort the training data to simulate image distortion caused by factors such as lens aging.

See also

set_enable()

Returns:

bool Enable

GeoAugDistortion &set_enable(bool enable)

Set Enable with bool value.

Enable to randomly distort the training data to simulate image distortion caused by factors such as lens aging.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugDistortion& the reference of this object.

const std::vector<double> &get_range() const

Get Distortion Strength Range.

Distortion Strength range, From 0 to 2, the distortion effect gradually increases

See also

set_range()

Returns:

const std::vector<double> & Distortion Strength Range

GeoAugDistortion &set_range(const std::vector<double> &range)

Set Distortion Strength Range with std::vector<double> value.

Distortion Strength range, From 0 to 2, the distortion effect gradually increases

See also

get_range()

Parameters:

range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugDistortion& the reference of this object.

bool range_contains(double value) const

Check if the Distortion Strength Range contains the value.

Returns:

bool true if the Distortion Strength Range contains the value, otherwise false.

double get_range_left() const

Get left point value of Distortion Strength Range.

See also

set_range_right()

Returns:

const std::vector<double> & left point value of Distortion Strength Range

double get_range_right() const

Get the right point value of Distortion Strength Range.

See also

set_range_left()

Returns:

const std::vector<double> & the right point value of Distortion Strength Range

GeoAugDistortion &set_range_left(double range_left)

Set left point value of Distortion Strength Range with std::vector<double> value.

Distortion Strength range, From 0 to 2, the distortion effect gradually increases

See also

get_range_left()

Parameters:

range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugDistortion& the reference of this object.

GeoAugDistortion &set_range_right(double range_right)

Set the right point value of Distortion Strength Range with std::vector<double> value.

Distortion Strength range, From 0 to 2, the distortion effect gradually increases

See also

get_range_right()

Parameters:

range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeoAugDistortion& the reference of this object.

class GeometryAugmentation : public visionflow::param::ISchemable

GeometryAugmentation Parameter class generated by jinja2 automatically.

Geometry augmentation parameters

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_flip_vertical() const

Get Vertical Flip.

Flip training images vertically randomly with a probability of 0.5

See also

set_flip_vertical()

Returns:

bool Vertical Flip

GeometryAugmentation &set_flip_vertical(bool flip_vertical)

Set Vertical Flip with bool value.

Flip training images vertically randomly with a probability of 0.5

See also

get_flip_vertical()

Parameters:

flip_vertical – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeometryAugmentation& the reference of this object.

bool get_flip_horizontal() const

Get Horizontal Flip.

Flip training images horizontally randomly with a probability of 0.5

See also

set_flip_horizontal()

Returns:

bool Horizontal Flip

GeometryAugmentation &set_flip_horizontal(bool flip_horizontal)

Set Horizontal Flip with bool value.

Flip training images horizontally randomly with a probability of 0.5

See also

get_flip_horizontal()

Parameters:

flip_horizontal – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeometryAugmentation& the reference of this object.

bool get_rotate_90() const

Get Vertical Rotation.

Rotate the training image by a multiple of 90 degrees at random

See also

set_rotate_90()

Returns:

bool Vertical Rotation

GeometryAugmentation &set_rotate_90(bool rotate_90)

Set Vertical Rotation with bool value.

Rotate the training image by a multiple of 90 degrees at random

See also

get_rotate_90()

Parameters:

rotate_90 – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeometryAugmentation& the reference of this object.

bool get_rotate_180() const

Get Centrosymmetric Rotation.

Rotate the training image by a multiple of 180 degrees at random

See also

set_rotate_180()

Returns:

bool Centrosymmetric Rotation

GeometryAugmentation &set_rotate_180(bool rotate_180)

Set Centrosymmetric Rotation with bool value.

Rotate the training image by a multiple of 180 degrees at random

See also

get_rotate_180()

Parameters:

rotate_180 – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeometryAugmentation& the reference of this object.

const GeoAugSlightRotation &get_slight_rotate() const

Get Slight Rotation.

Slight rotation augmentation parameters

See also

set_slight_rotate()

Returns:

const GeoAugSlightRotation & Slight Rotation

GeometryAugmentation &set_slight_rotate(const GeoAugSlightRotation &slight_rotate)

Set Slight Rotation with GeoAugSlightRotation value.

Slight rotation augmentation parameters

See also

get_slight_rotate()

Parameters:

slight_rotate – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeometryAugmentation& the reference of this object.

GeoAugSlightRotation &get_slight_rotate()

Get mutable reference of Slight Rotation.

Slight rotation augmentation parameters

Returns:

GeometryAugmentation& the mutable reference of the group.

const GeoAugShift &get_shift() const

Get Shift.

Shift augmentations parameters

See also

set_shift()

Returns:

const GeoAugShift & Shift

GeometryAugmentation &set_shift(const GeoAugShift &shift)

Set Shift with GeoAugShift value.

Shift augmentations parameters

See also

get_shift()

Parameters:

shift – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeometryAugmentation& the reference of this object.

GeoAugShift &get_shift()

Get mutable reference of Shift.

Shift augmentations parameters

Returns:

GeometryAugmentation& the mutable reference of the group.

const GeoAugScale &get_scale() const

Get Scale.

Scale augmentations parameters

See also

set_scale()

Returns:

const GeoAugScale & Scale

GeometryAugmentation &set_scale(const GeoAugScale &scale)

Set Scale with GeoAugScale value.

Scale augmentations parameters

See also

get_scale()

Parameters:

scale – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeometryAugmentation& the reference of this object.

GeoAugScale &get_scale()

Get mutable reference of Scale.

Scale augmentations parameters

Returns:

GeometryAugmentation& the mutable reference of the group.

const GeoAugDistortion &get_distortion() const

Get Distortion.

Distortion augmentations parameters

See also

set_distortion()

Returns:

const GeoAugDistortion & Distortion

GeometryAugmentation &set_distortion(const GeoAugDistortion &distortion)

Set Distortion with GeoAugDistortion value.

Distortion augmentations parameters

See also

get_distortion()

Parameters:

distortion – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeometryAugmentation& the reference of this object.

GeoAugDistortion &get_distortion()

Get mutable reference of Distortion.

Distortion augmentations parameters

Returns:

GeometryAugmentation& the mutable reference of the group.

bool get_crop() const

Get Crop Overflow Area.

Crop out spill areas caused by geometric transformations (translation, rotation) to keep the image size unchanged

See also

set_crop()

Returns:

bool Crop Overflow Area

GeometryAugmentation &set_crop(bool crop)

Set Crop Overflow Area with bool value.

Crop out spill areas caused by geometric transformations (translation, rotation) to keep the image size unchanged

See also

get_crop()

Parameters:

crop – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

GeometryAugmentation& the reference of this object.

class ImageAugBrightness : public visionflow::param::ISchemable

ImageAugBrightness Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Brightness Augmentation.

Enable illumination augmentation to augment training data with linear gray scale transformations

See also

set_enable()

Returns:

bool Enable Brightness Augmentation

ImageAugBrightness &set_enable(bool enable)

Set Enable Brightness Augmentation with bool value.

Enable illumination augmentation to augment training data with linear gray scale transformations

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugBrightness& the reference of this object.

const std::vector<double> &get_range() const

Get Brightness Range.

The illumination intensity range 0~1 for reduction, and 1~2 for enhancement.

See also

set_range()

Returns:

const std::vector<double> & Brightness Range

ImageAugBrightness &set_range(const std::vector<double> &range)

Set Brightness Range with std::vector<double> value.

The illumination intensity range 0~1 for reduction, and 1~2 for enhancement.

See also

get_range()

Parameters:

range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugBrightness& the reference of this object.

bool range_contains(double value) const

Check if the Brightness Range contains the value.

Returns:

bool true if the Brightness Range contains the value, otherwise false.

double get_range_left() const

Get left point value of Brightness Range.

See also

set_range_right()

Returns:

const std::vector<double> & left point value of Brightness Range

double get_range_right() const

Get the right point value of Brightness Range.

See also

set_range_left()

Returns:

const std::vector<double> & the right point value of Brightness Range

ImageAugBrightness &set_range_left(double range_left)

Set left point value of Brightness Range with std::vector<double> value.

The illumination intensity range 0~1 for reduction, and 1~2 for enhancement.

See also

get_range_left()

Parameters:

range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugBrightness& the reference of this object.

ImageAugBrightness &set_range_right(double range_right)

Set the right point value of Brightness Range with std::vector<double> value.

The illumination intensity range 0~1 for reduction, and 1~2 for enhancement.

See also

get_range_right()

Parameters:

range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugBrightness& the reference of this object.

class ImageAugContrast : public visionflow::param::ISchemable

ImageAugContrast Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Contrast Augmentation.

Enable contrast augmentation to simulate the effect of different lighting conditions with different contrast

See also

set_enable()

Returns:

bool Enable Contrast Augmentation

ImageAugContrast &set_enable(bool enable)

Set Enable Contrast Augmentation with bool value.

Enable contrast augmentation to simulate the effect of different lighting conditions with different contrast

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugContrast& the reference of this object.

const std::vector<double> &get_range() const

Get Contrast Range.

The contrast range 0~1 for reduction, and 1~2 for enhancement.

See also

set_range()

Returns:

const std::vector<double> & Contrast Range

ImageAugContrast &set_range(const std::vector<double> &range)

Set Contrast Range with std::vector<double> value.

The contrast range 0~1 for reduction, and 1~2 for enhancement.

See also

get_range()

Parameters:

range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugContrast& the reference of this object.

bool range_contains(double value) const

Check if the Contrast Range contains the value.

Returns:

bool true if the Contrast Range contains the value, otherwise false.

double get_range_left() const

Get left point value of Contrast Range.

See also

set_range_right()

Returns:

const std::vector<double> & left point value of Contrast Range

double get_range_right() const

Get the right point value of Contrast Range.

See also

set_range_left()

Returns:

const std::vector<double> & the right point value of Contrast Range

ImageAugContrast &set_range_left(double range_left)

Set left point value of Contrast Range with std::vector<double> value.

The contrast range 0~1 for reduction, and 1~2 for enhancement.

See also

get_range_left()

Parameters:

range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugContrast& the reference of this object.

ImageAugContrast &set_range_right(double range_right)

Set the right point value of Contrast Range with std::vector<double> value.

The contrast range 0~1 for reduction, and 1~2 for enhancement.

See also

get_range_right()

Parameters:

range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugContrast& the reference of this object.

class ImageAugColorFilter : public visionflow::param::ISchemable

ImageAugColorFilter Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Color Filter.

Simulate the different colors lighting effects by adding filters (only color images are supported), the color filter intensity controls the max intensity of the allowed color filters, range 0~2 from low to high.

See also

set_enable()

Returns:

bool Enable Color Filter

ImageAugColorFilter &set_enable(bool enable)

Set Enable Color Filter with bool value.

Simulate the different colors lighting effects by adding filters (only color images are supported), the color filter intensity controls the max intensity of the allowed color filters, range 0~2 from low to high.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugColorFilter& the reference of this object.

const std::vector<double> &get_range() const

Get Color Filter Range.

Color Filter Range 0~2 from low to high.

See also

set_range()

Returns:

const std::vector<double> & Color Filter Range

ImageAugColorFilter &set_range(const std::vector<double> &range)

Set Color Filter Range with std::vector<double> value.

Color Filter Range 0~2 from low to high.

See also

get_range()

Parameters:

range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugColorFilter& the reference of this object.

bool range_contains(double value) const

Check if the Color Filter Range contains the value.

Returns:

bool true if the Color Filter Range contains the value, otherwise false.

double get_range_left() const

Get left point value of Color Filter Range.

See also

set_range_right()

Returns:

const std::vector<double> & left point value of Color Filter Range

double get_range_right() const

Get the right point value of Color Filter Range.

See also

set_range_left()

Returns:

const std::vector<double> & the right point value of Color Filter Range

ImageAugColorFilter &set_range_left(double range_left)

Set left point value of Color Filter Range with std::vector<double> value.

Color Filter Range 0~2 from low to high.

See also

get_range_left()

Parameters:

range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugColorFilter& the reference of this object.

ImageAugColorFilter &set_range_right(double range_right)

Set the right point value of Color Filter Range with std::vector<double> value.

Color Filter Range 0~2 from low to high.

See also

get_range_right()

Parameters:

range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugColorFilter& the reference of this object.

class ImageAugNoise : public visionflow::param::ISchemable

ImageAugNoise Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Noise.

Enable noise augmentation to augment training data with Gaussian noise

See also

set_enable()

Returns:

bool Enable Noise

ImageAugNoise &set_enable(bool enable)

Set Enable Noise with bool value.

Enable noise augmentation to augment training data with Gaussian noise

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugNoise& the reference of this object.

const std::vector<double> &get_range() const

Get Noise Range.

Noise Range 0~2 from low to high.

See also

set_range()

Returns:

const std::vector<double> & Noise Range

ImageAugNoise &set_range(const std::vector<double> &range)

Set Noise Range with std::vector<double> value.

Noise Range 0~2 from low to high.

See also

get_range()

Parameters:

range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugNoise& the reference of this object.

bool range_contains(double value) const

Check if the Noise Range contains the value.

Returns:

bool true if the Noise Range contains the value, otherwise false.

double get_range_left() const

Get left point value of Noise Range.

See also

set_range_right()

Returns:

const std::vector<double> & left point value of Noise Range

double get_range_right() const

Get the right point value of Noise Range.

See also

set_range_left()

Returns:

const std::vector<double> & the right point value of Noise Range

ImageAugNoise &set_range_left(double range_left)

Set left point value of Noise Range with std::vector<double> value.

Noise Range 0~2 from low to high.

See also

get_range_left()

Parameters:

range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugNoise& the reference of this object.

ImageAugNoise &set_range_right(double range_right)

Set the right point value of Noise Range with std::vector<double> value.

Noise Range 0~2 from low to high.

See also

get_range_right()

Parameters:

range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugNoise& the reference of this object.

class ImageAugIlluminationGradient : public visionflow::param::ISchemable

ImageAugIlluminationGradient Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Illumination Gradient.

Simulate the illumination intensity changing gradient caused by the shift of the light position, the illumination intensity range 0~2 from low to high

See also

set_enable()

Returns:

bool Enable Illumination Gradient

ImageAugIlluminationGradient &set_enable(bool enable)

Set Enable Illumination Gradient with bool value.

Simulate the illumination intensity changing gradient caused by the shift of the light position, the illumination intensity range 0~2 from low to high

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugIlluminationGradient& the reference of this object.

const std::vector<double> &get_range() const

Get Illumination Gradient Range.

Illumination Gradient Range 0~2 from low to high.

See also

set_range()

Returns:

const std::vector<double> & Illumination Gradient Range

ImageAugIlluminationGradient &set_range(const std::vector<double> &range)

Set Illumination Gradient Range with std::vector<double> value.

Illumination Gradient Range 0~2 from low to high.

See also

get_range()

Parameters:

range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugIlluminationGradient& the reference of this object.

bool range_contains(double value) const

Check if the Illumination Gradient Range contains the value.

Returns:

bool true if the Illumination Gradient Range contains the value, otherwise false.

double get_range_left() const

Get left point value of Illumination Gradient Range.

See also

set_range_right()

Returns:

const std::vector<double> & left point value of Illumination Gradient Range

double get_range_right() const

Get the right point value of Illumination Gradient Range.

See also

set_range_left()

Returns:

const std::vector<double> & the right point value of Illumination Gradient Range

ImageAugIlluminationGradient &set_range_left(double range_left)

Set left point value of Illumination Gradient Range with std::vector<double> value.

Illumination Gradient Range 0~2 from low to high.

See also

get_range_left()

Parameters:

range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugIlluminationGradient& the reference of this object.

ImageAugIlluminationGradient &set_range_right(double range_right)

Set the right point value of Illumination Gradient Range with std::vector<double> value.

Illumination Gradient Range 0~2 from low to high.

See also

get_range_right()

Parameters:

range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugIlluminationGradient& the reference of this object.

class ImageAugSmoothingOrSharpening : public visionflow::param::ISchemable

ImageAugSmoothingOrSharpening Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Smoothing/Sharpening.

Simulate a scene with more accurate lens focus by sharpening the image, sharpening intensity range 0~2 from low to high

See also

set_enable()

Returns:

bool Enable Smoothing/Sharpening

ImageAugSmoothingOrSharpening &set_enable(bool enable)

Set Enable Smoothing/Sharpening with bool value.

Simulate a scene with more accurate lens focus by sharpening the image, sharpening intensity range 0~2 from low to high

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugSmoothingOrSharpening& the reference of this object.

const std::vector<double> &get_range() const

Get Smoothing/Sharpening Range.

Randomly smooth or sharpen training images, where -1 to 0 indicates smoothing and 0 to 1 indicates sharpening.

See also

set_range()

Returns:

const std::vector<double> & Smoothing/Sharpening Range

ImageAugSmoothingOrSharpening &set_range(const std::vector<double> &range)

Set Smoothing/Sharpening Range with std::vector<double> value.

Randomly smooth or sharpen training images, where -1 to 0 indicates smoothing and 0 to 1 indicates sharpening.

See also

get_range()

Parameters:

range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugSmoothingOrSharpening& the reference of this object.

bool range_contains(double value) const

Check if the Smoothing/Sharpening Range contains the value.

Returns:

bool true if the Smoothing/Sharpening Range contains the value, otherwise false.

double get_range_left() const

Get left point value of Smoothing/Sharpening Range.

See also

set_range_right()

Returns:

const std::vector<double> & left point value of Smoothing/Sharpening Range

double get_range_right() const

Get the right point value of Smoothing/Sharpening Range.

See also

set_range_left()

Returns:

const std::vector<double> & the right point value of Smoothing/Sharpening Range

ImageAugSmoothingOrSharpening &set_range_left(double range_left)

Set left point value of Smoothing/Sharpening Range with std::vector<double> value.

Randomly smooth or sharpen training images, where -1 to 0 indicates smoothing and 0 to 1 indicates sharpening.

See also

get_range_left()

Parameters:

range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugSmoothingOrSharpening& the reference of this object.

ImageAugSmoothingOrSharpening &set_range_right(double range_right)

Set the right point value of Smoothing/Sharpening Range with std::vector<double> value.

Randomly smooth or sharpen training images, where -1 to 0 indicates smoothing and 0 to 1 indicates sharpening.

See also

get_range_right()

Parameters:

range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugSmoothingOrSharpening& the reference of this object.

class ImageAugmentation : public visionflow::param::ISchemable

ImageAugmentation Parameter class generated by jinja2 automatically.

Image Augmentation Parameters Group.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const ImageAugBrightness &get_brightness() const

Get Illumination Augmentation.

Illumination augmentation parameters

See also

set_brightness()

Returns:

const ImageAugBrightness & Illumination Augmentation

ImageAugmentation &set_brightness(const ImageAugBrightness &brightness)

Set Illumination Augmentation with ImageAugBrightness value.

Illumination augmentation parameters

See also

get_brightness()

Parameters:

brightness – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugmentation& the reference of this object.

ImageAugBrightness &get_brightness()

Get mutable reference of Illumination Augmentation.

Illumination augmentation parameters

Returns:

ImageAugmentation& the mutable reference of the group.

const ImageAugContrast &get_contrast() const

Get Contrast Augmentation.

Contrast augmentation parameters

See also

set_contrast()

Returns:

const ImageAugContrast & Contrast Augmentation

ImageAugmentation &set_contrast(const ImageAugContrast &contrast)

Set Contrast Augmentation with ImageAugContrast value.

Contrast augmentation parameters

See also

get_contrast()

Parameters:

contrast – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugmentation& the reference of this object.

ImageAugContrast &get_contrast()

Get mutable reference of Contrast Augmentation.

Contrast augmentation parameters

Returns:

ImageAugmentation& the mutable reference of the group.

const ImageAugNoise &get_noise() const

Get Noise Augmentation.

Noise augmentation parameters

See also

set_noise()

Returns:

const ImageAugNoise & Noise Augmentation

ImageAugmentation &set_noise(const ImageAugNoise &noise)

Set Noise Augmentation with ImageAugNoise value.

Noise augmentation parameters

See also

get_noise()

Parameters:

noise – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugmentation& the reference of this object.

ImageAugNoise &get_noise()

Get mutable reference of Noise Augmentation.

Noise augmentation parameters

Returns:

ImageAugmentation& the mutable reference of the group.

const ImageAugSmoothingOrSharpening &get_blur() const

Get Smoothing/Sharpening.

Blur augmentation parameters

See also

set_blur()

Returns:

const ImageAugSmoothingOrSharpening & Smoothing/Sharpening

ImageAugmentation &set_blur(const ImageAugSmoothingOrSharpening &blur)

Set Smoothing/Sharpening with ImageAugSmoothingOrSharpening value.

Blur augmentation parameters

See also

get_blur()

Parameters:

blur – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugmentation& the reference of this object.

ImageAugSmoothingOrSharpening &get_blur()

Get mutable reference of Smoothing/Sharpening.

Blur augmentation parameters

Returns:

ImageAugmentation& the mutable reference of the group.

const ImageAugColorFilter &get_color_filter() const

Get Color Filter.

Color filter augmentation parameters

See also

set_color_filter()

Returns:

const ImageAugColorFilter & Color Filter

ImageAugmentation &set_color_filter(const ImageAugColorFilter &color_filter)

Set Color Filter with ImageAugColorFilter value.

Color filter augmentation parameters

See also

get_color_filter()

Parameters:

color_filter – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugmentation& the reference of this object.

ImageAugColorFilter &get_color_filter()

Get mutable reference of Color Filter.

Color filter augmentation parameters

Returns:

ImageAugmentation& the mutable reference of the group.

const ImageAugIlluminationGradient &get_illumination_gradient() const

Get Illumination Gradient.

Illumination gradient augmentation parameters

See also

set_illumination_gradient()

Returns:

const ImageAugIlluminationGradient & Illumination Gradient

ImageAugmentation &set_illumination_gradient(const ImageAugIlluminationGradient &illumination_gradient)

Set Illumination Gradient with ImageAugIlluminationGradient value.

Illumination gradient augmentation parameters

See also

get_illumination_gradient()

Parameters:

illumination_gradient – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ImageAugmentation& the reference of this object.

ImageAugIlluminationGradient &get_illumination_gradient()

Get mutable reference of Illumination Gradient.

Illumination gradient augmentation parameters

Returns:

ImageAugmentation& the mutable reference of the group.

class DataAugmentation : public visionflow::param::ISchemable

DataAugmentation Parameter class generated by jinja2 automatically.

Data Augmentation Parameters Group.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const GeometryAugmentation &get_geometry_augmentation() const

Get Geometry Augmentation.

Geometry augmentation parameters

See also

set_geometry_augmentation()

Returns:

const GeometryAugmentation & Geometry Augmentation

DataAugmentation &set_geometry_augmentation(const GeometryAugmentation &geometry_augmentation)

Set Geometry Augmentation with GeometryAugmentation value.

Geometry augmentation parameters

See also

get_geometry_augmentation()

Parameters:

geometry_augmentation – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DataAugmentation& the reference of this object.

GeometryAugmentation &get_geometry_augmentation()

Get mutable reference of Geometry Augmentation.

Geometry augmentation parameters

Returns:

DataAugmentation& the mutable reference of the group.

const ImageAugmentation &get_image_augmentation() const

Get Image Augmentation.

Image augmentation parameters

See also

set_image_augmentation()

Returns:

const ImageAugmentation & Image Augmentation

DataAugmentation &set_image_augmentation(const ImageAugmentation &image_augmentation)

Set Image Augmentation with ImageAugmentation value.

Image augmentation parameters

See also

get_image_augmentation()

Parameters:

image_augmentation – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DataAugmentation& the reference of this object.

ImageAugmentation &get_image_augmentation()

Get mutable reference of Image Augmentation.

Image augmentation parameters

Returns:

DataAugmentation& the mutable reference of the group.

Segmentation Training Parameters

class SegmentationInputShape : public visionflow::param::ISchemable

SegmentationInputShape Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Custom Input Shape.

Enable to Customize Network Input Shape

See also

set_enable()

Returns:

bool Enable Custom Input Shape

SegmentationInputShape &set_enable(bool enable)

Set Enable Custom Input Shape with bool value.

Enable to Customize Network Input Shape

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationInputShape& the reference of this object.

int get_base_input_width() const

Get Base Input Width.

Base Input Width

See also

set_base_input_width()

Returns:

int Base Input Width

SegmentationInputShape &set_base_input_width(int base_input_width)

Set Base Input Width with int value.

Base Input Width

See also

get_base_input_width()

Parameters:

base_input_width – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationInputShape& the reference of this object.

int get_base_input_height() const

Get Base Input Height.

Base Input Height

See also

set_base_input_height()

Returns:

int Base Input Height

SegmentationInputShape &set_base_input_height(int base_input_height)

Set Base Input Height with int value.

Base Input Height

See also

get_base_input_height()

Parameters:

base_input_height – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationInputShape& the reference of this object.

enum visionflow::param::ImageSplitMode

Values:

enumerator kAuto = 0

enumerator kManual = 1

class SegmentationImageSplit : public visionflow::param::ISchemable

SegmentationImageSplit Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const ImageSplitMode &get_split_set_method() const

Get Split Method.

kAuto : The algorithm adapts the recommended resolution levelkManual according to the original map size and minimum marking size: the resolution level is based on manual setting view : The actual resolution level can be viewed in Help-View log. Br/>Brief description of algorithm : The algorithm scales the short edge of the image to 256*resolution level before feeding to the network, and the long edge scales equally to the input network.

See also

set_split_set_method()

Returns:

const ImageSplitMode & Split Method

SegmentationImageSplit &set_split_set_method(const ImageSplitMode &split_set_method)

Set Split Method with ImageSplitMode value.

kAuto : The algorithm adapts the recommended resolution levelkManual according to the original map size and minimum marking size: the resolution level is based on manual setting view : The actual resolution level can be viewed in Help-View log. Br/>Brief description of algorithm : The algorithm scales the short edge of the image to 256*resolution level before feeding to the network, and the long edge scales equally to the input network.

See also

get_split_set_method()

Parameters:

split_set_method – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationImageSplit& the reference of this object.

int get_max_split() const

Get Resolution Level.

The short side of the image input will be scaled to (base_size*Resolution Level), and the long side will be scaled proportionally. When the max precision set is higher than the required precision of the algorithm, the precision will be automatically reduced. The applied precision can be acquired from figuration files of the trained model

See also

set_max_split()

Returns:

int Resolution Level

SegmentationImageSplit &set_max_split(int max_split)

Set Resolution Level with int value.

The short side of the image input will be scaled to (base_size*Resolution Level), and the long side will be scaled proportionally. When the max precision set is higher than the required precision of the algorithm, the precision will be automatically reduced. The applied precision can be acquired from figuration files of the trained model

See also

get_max_split()

Parameters:

max_split – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationImageSplit& the reference of this object.

class SegmentationModelParameters : public visionflow::param::ISchemable

SegmentationModelParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const std::string &get_model_arch() const

Get Model Architecture.

The ‘Small Defects’ model has better detection effect on small defects, and the ‘Comprehensive Model’ model has better comprehensive effect. Recommended to use the ‘Comprehensive Model’ model when the detection effect of the Small Defects model is poor. ‘Contrastive Model’ is only for contrastive segmentation

See also

set_model_arch()

Returns:

const std::string & Model Architecture

SegmentationModelParameters &set_model_arch(const std::string &model_arch)

Set Model Architecture with std::string value.

The ‘Small Defects’ model has better detection effect on small defects, and the ‘Comprehensive Model’ model has better comprehensive effect. Recommended to use the ‘Comprehensive Model’ model when the detection effect of the Small Defects model is poor. ‘Contrastive Model’ is only for contrastive segmentation

See also

get_model_arch()

Parameters:

model_arch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationModelParameters& the reference of this object.

bool get_shift_stable() const

Get With Shift Stable.

With shift stable, the adaptability of the model to the slight shift of the target can be improved, but the training and inference speed will be 20 percent slower

See also

set_shift_stable()

Returns:

bool With Shift Stable

SegmentationModelParameters &set_shift_stable(bool shift_stable)

Set With Shift Stable with bool value.

With shift stable, the adaptability of the model to the slight shift of the target can be improved, but the training and inference speed will be 20 percent slower

See also

get_shift_stable()

Parameters:

shift_stable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationModelParameters& the reference of this object.

class SegmentationTrainingSampleStrategy : public visionflow::param::ISchemable

SegmentationTrainingSampleStrategy Parameter class generated by jinja2 automatically.

Segmentation training parameters to control the the input data.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

double get_hard_case_rate() const

Get Sample Rate of Hard Case.

The larger the value set, the higher the attention to the hard case

See also

set_hard_case_rate()

Returns:

double Sample Rate of Hard Case

SegmentationTrainingSampleStrategy &set_hard_case_rate(double hard_case_rate)

Set Sample Rate of Hard Case with double value.

The larger the value set, the higher the attention to the hard case

See also

get_hard_case_rate()

Parameters:

hard_case_rate – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationTrainingSampleStrategy& the reference of this object.

int get_hard_case_split() const

Get Hard Case Split.

Number of hard case discrete coordinates, Default value is recommended.

See also

set_hard_case_split()

Returns:

int Hard Case Split

SegmentationTrainingSampleStrategy &set_hard_case_split(int hard_case_split)

Set Hard Case Split with int value.

Number of hard case discrete coordinates, Default value is recommended.

See also

get_hard_case_split()

Parameters:

hard_case_split – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationTrainingSampleStrategy& the reference of this object.

const SegmentationImageSplit &get_image_split() const

Get Image Split.

Image Split Parameters

See also

set_image_split()

Returns:

const SegmentationImageSplit & Image Split

SegmentationTrainingSampleStrategy &set_image_split(const SegmentationImageSplit &image_split)

Set Image Split with SegmentationImageSplit value.

Image Split Parameters

See also

get_image_split()

Parameters:

image_split – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationTrainingSampleStrategy& the reference of this object.

SegmentationImageSplit &get_image_split()

Get mutable reference of Image Split.

Image Split Parameters

Returns:

SegmentationTrainingSampleStrategy& the mutable reference of the group.

const SegmentationInputShape &get_input_shape() const

Get Input Shape.

Input Shape Parameters

See also

set_input_shape()

Returns:

const SegmentationInputShape & Input Shape

SegmentationTrainingSampleStrategy &set_input_shape(const SegmentationInputShape &input_shape)

Set Input Shape with SegmentationInputShape value.

Input Shape Parameters

See also

get_input_shape()

Parameters:

input_shape – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationTrainingSampleStrategy& the reference of this object.

SegmentationInputShape &get_input_shape()

Get mutable reference of Input Shape.

Input Shape Parameters

Returns:

SegmentationTrainingSampleStrategy& the mutable reference of the group.

class SegmentationTrainingParameters : public visionflow::param::SchemableParameter

SegmentationTrainingParameters Parameter class generated by jinja2 automatically.

Segmentation Training Parameters Group.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int get_epoch() const

Get Epoch.

Number of epochs to train

See also

set_epoch()

Returns:

int Epoch

SegmentationTrainingParameters &set_epoch(int epoch)

Set Epoch with int value.

Number of epochs to train

See also

get_epoch()

Parameters:

epoch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationTrainingParameters& the reference of this object.

int get_batch_size() const

Get Training Batch Size.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16. The classification module needs to be set larger, generally 32, 64

See also

set_batch_size()

Returns:

int Training Batch Size

SegmentationTrainingParameters &set_batch_size(int batch_size)

Set Training Batch Size with int value.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16. The classification module needs to be set larger, generally 32, 64

See also

get_batch_size()

Parameters:

batch_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationTrainingParameters& the reference of this object.

const SegmentationTrainingSampleStrategy &get_sampling_strategy() const

Get Sampling Strategy.

Sampling Strategy Parameters

See also

set_sampling_strategy()

Returns:

const SegmentationTrainingSampleStrategy & Sampling Strategy

SegmentationTrainingParameters &set_sampling_strategy(const SegmentationTrainingSampleStrategy &sampling_strategy)

Set Sampling Strategy with SegmentationTrainingSampleStrategy value.

Sampling Strategy Parameters

See also

get_sampling_strategy()

Parameters:

sampling_strategy – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationTrainingParameters& the reference of this object.

SegmentationTrainingSampleStrategy &get_sampling_strategy()

Get mutable reference of Sampling Strategy.

Sampling Strategy Parameters

Returns:

SegmentationTrainingParameters& the mutable reference of the group.

const TrainingMode &get_training_mode() const

Get Training Mode.

The ‘Normal’ mode is the default mode, which trains the model from scratch/pre-train. The ‘Incremental’ mode uses an incremental training algorithm to update the model parameters based on new data, without retraining the entire model from scratch.

See also

set_training_mode()

Returns:

const TrainingMode & Training Mode

SegmentationTrainingParameters &set_training_mode(const TrainingMode &training_mode)

Set Training Mode with TrainingMode value.

The ‘Normal’ mode is the default mode, which trains the model from scratch/pre-train. The ‘Incremental’ mode uses an incremental training algorithm to update the model parameters based on new data, without retraining the entire model from scratch.

See also

get_training_mode()

Parameters:

training_mode – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationTrainingParameters& the reference of this object.

bool get_only_check() const

Get Only Check.

Default is False. If set to True, the trainer will only check if it is possible to train in the given training mode without actually performing the training.

See also

set_only_check()

Returns:

bool Only Check

SegmentationTrainingParameters &set_only_check(bool only_check)

Set Only Check with bool value.

Default is False. If set to True, the trainer will only check if it is possible to train in the given training mode without actually performing the training.

See also

get_only_check()

Parameters:

only_check – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationTrainingParameters& the reference of this object.

const SegmentationModelParameters &get_model_param() const

Get Model Parameters.

Select the appropriate basic model according to different detection targets and speed requirements

See also

set_model_param()

Returns:

const SegmentationModelParameters & Model Parameters

SegmentationTrainingParameters &set_model_param(const SegmentationModelParameters &model_param)

Set Model Parameters with SegmentationModelParameters value.

Select the appropriate basic model according to different detection targets and speed requirements

See also

get_model_param()

Parameters:

model_param – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationTrainingParameters& the reference of this object.

SegmentationModelParameters &get_model_param()

Get mutable reference of Model Parameters.

Select the appropriate basic model according to different detection targets and speed requirements

Returns:

SegmentationTrainingParameters& the mutable reference of the group.

const DataAugmentation &get_augmentations() const

Get Data Augmentation.

Data Augmentation Parameters

See also

set_augmentations()

Returns:

const DataAugmentation & Data Augmentation

SegmentationTrainingParameters &set_augmentations(const DataAugmentation &augmentations)

Set Data Augmentation with DataAugmentation value.

Data Augmentation Parameters

See also

get_augmentations()

Parameters:

augmentations – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SegmentationTrainingParameters& the reference of this object.

DataAugmentation &get_augmentations()

Get mutable reference of Data Augmentation.

Data Augmentation Parameters

Returns:

SegmentationTrainingParameters& the mutable reference of the group.

class BinaryTrainingParameters : public visionflow::param::IParameter

Parameter class to store some binary configuration information that the parameter generator can not represent for the model trainer.

Public Functions

BinaryTrainingParameters &set_global_mask(const geometry::MultiPolygon2f &masks)

Set the global mask. Each sub image in the training views will be masked with the image mean values before training.

Parameters:

masks – The mask polygons.

Returns:

BinaryTrainingParameters& The parameter instance.

const geometry::MultiPolygon2f &get_global_mask() const

Get the global mask.

Returns:

const geometry::MultiPolygon2f& The global mask polygons

Unsuper Segmentation Training Parameters

class UnsuperSegmentationInputShape : public visionflow::param::ISchemable

UnsuperSegmentationInputShape Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Custom Sampling Shape.

Enable to Customize Network Sampling Input Shape

See also

set_enable()

Returns:

bool Enable Custom Sampling Shape

UnsuperSegmentationInputShape &set_enable(bool enable)

Set Enable Custom Sampling Shape with bool value.

Enable to Customize Network Sampling Input Shape

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationInputShape& the reference of this object.

int get_sampling_width() const

Get Sampling Input Width.

Sampling Input Width

See also

set_sampling_width()

Returns:

int Sampling Input Width

UnsuperSegmentationInputShape &set_sampling_width(int sampling_width)

Set Sampling Input Width with int value.

Sampling Input Width

See also

get_sampling_width()

Parameters:

sampling_width – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationInputShape& the reference of this object.

int get_sampling_height() const

Get Sampling Input Height.

Sampling Input Height

See also

set_sampling_height()

Returns:

int Sampling Input Height

UnsuperSegmentationInputShape &set_sampling_height(int sampling_height)

Set Sampling Input Height with int value.

Sampling Input Height

See also

get_sampling_height()

Parameters:

sampling_height – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationInputShape& the reference of this object.

class UnsuperSegmentationTrainingParameters : public visionflow::param::SchemableParameter

UnsuperSegmentationTrainingParameters Parameter class generated by jinja2 automatically.

Unsuper Segmentation Training Parameters Group.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int get_epoch() const

Get Epochs.

Number of epochs to train

See also

set_epoch()

Returns:

int Epochs

UnsuperSegmentationTrainingParameters &set_epoch(int epoch)

Set Epochs with int value.

Number of epochs to train

See also

get_epoch()

Parameters:

epoch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationTrainingParameters& the reference of this object.

int get_batch_size() const

Get Batch Size.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16. The classification module needs to be set larger, generally 32, 64

See also

set_batch_size()

Returns:

int Batch Size

UnsuperSegmentationTrainingParameters &set_batch_size(int batch_size)

Set Batch Size with int value.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16. The classification module needs to be set larger, generally 32, 64

See also

get_batch_size()

Parameters:

batch_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationTrainingParameters& the reference of this object.

const UnsuperSegmentationSamplingParameters &get_sampling_strategy() const

Get Sampling Strategy.

Sampling Parameters

See also

set_sampling_strategy()

Returns:

const UnsuperSegmentationSamplingParameters & Sampling Strategy

UnsuperSegmentationTrainingParameters &set_sampling_strategy(const UnsuperSegmentationSamplingParameters &sampling_strategy)

Set Sampling Strategy with UnsuperSegmentationSamplingParameters value.

Sampling Parameters

See also

get_sampling_strategy()

Parameters:

sampling_strategy – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationTrainingParameters& the reference of this object.

UnsuperSegmentationSamplingParameters &get_sampling_strategy()

Get mutable reference of Sampling Strategy.

Sampling Parameters

Returns:

UnsuperSegmentationTrainingParameters& the mutable reference of the group.

const UnsuperSegmentationModelParameters &get_model_param() const

Get Model Parameters.

Model Parameters

See also

set_model_param()

Returns:

const UnsuperSegmentationModelParameters & Model Parameters

UnsuperSegmentationTrainingParameters &set_model_param(const UnsuperSegmentationModelParameters &model_param)

Set Model Parameters with UnsuperSegmentationModelParameters value.

Model Parameters

See also

get_model_param()

Parameters:

model_param – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationTrainingParameters& the reference of this object.

UnsuperSegmentationModelParameters &get_model_param()

Get mutable reference of Model Parameters.

Model Parameters

Returns:

UnsuperSegmentationTrainingParameters& the mutable reference of the group.

const DataAugmentation &get_augmentations() const

Get Data Augmentation.

Data Augmentation Parameters

See also

set_augmentations()

Returns:

const DataAugmentation & Data Augmentation

UnsuperSegmentationTrainingParameters &set_augmentations(const DataAugmentation &augmentations)

Set Data Augmentation with DataAugmentation value.

Data Augmentation Parameters

See also

get_augmentations()

Parameters:

augmentations – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationTrainingParameters& the reference of this object.

DataAugmentation &get_augmentations()

Get mutable reference of Data Augmentation.

Data Augmentation Parameters

Returns:

UnsuperSegmentationTrainingParameters& the mutable reference of the group.

class UnsuperSegmentationSamplingParameters : public visionflow::param::ISchemable

UnsuperSegmentationSamplingParameters Parameter class generated by jinja2 automatically.

Unsuper Segmentation Training Sample Parameters Group.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const UnsuperSegmentationInputShape &get_sampling_input_shape() const

Get Input Shape.

Input Sampling Shape Parameters

See also

set_sampling_input_shape()

Returns:

const UnsuperSegmentationInputShape & Input Shape

UnsuperSegmentationSamplingParameters &set_sampling_input_shape(const UnsuperSegmentationInputShape &sampling_input_shape)

Set Input Shape with UnsuperSegmentationInputShape value.

Input Sampling Shape Parameters

See also

get_sampling_input_shape()

Parameters:

sampling_input_shape – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationSamplingParameters& the reference of this object.

UnsuperSegmentationInputShape &get_sampling_input_shape()

Get mutable reference of Input Shape.

Input Sampling Shape Parameters

Returns:

UnsuperSegmentationSamplingParameters& the mutable reference of the group.

bool get_local_sampling() const

Get Local Sampling Strategy.

when enable : Crop a patch(Input Shape) of image to train instead of whole image, increasing train speed but might make performance decline when image is complex

See also

set_local_sampling()

Returns:

bool Local Sampling Strategy

UnsuperSegmentationSamplingParameters &set_local_sampling(bool local_sampling)

Set Local Sampling Strategy with bool value.

when enable : Crop a patch(Input Shape) of image to train instead of whole image, increasing train speed but might make performance decline when image is complex

See also

get_local_sampling()

Parameters:

local_sampling – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationSamplingParameters& the reference of this object.

int get_sampling_resolution() const

Get Sampling Resolution Level.

The short side of the image input will be scaled to (Sampling Input Size * Sampling Resolution), and the long side will be scaled proportionally.

See also

set_sampling_resolution()

Returns:

int Sampling Resolution Level

UnsuperSegmentationSamplingParameters &set_sampling_resolution(int sampling_resolution)

Set Sampling Resolution Level with int value.

The short side of the image input will be scaled to (Sampling Input Size * Sampling Resolution), and the long side will be scaled proportionally.

See also

get_sampling_resolution()

Parameters:

sampling_resolution – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationSamplingParameters& the reference of this object.

bool get_image_consistent_optimize() const

Get High Consistent Image.

enable when image looks like the same in all region, such as texture image, it helps speed up train and might get better performance in such kind of image

See also

set_image_consistent_optimize()

Returns:

bool High Consistent Image

UnsuperSegmentationSamplingParameters &set_image_consistent_optimize(bool image_consistent_optimize)

Set High Consistent Image with bool value.

enable when image looks like the same in all region, such as texture image, it helps speed up train and might get better performance in such kind of image

See also

get_image_consistent_optimize()

Parameters:

image_consistent_optimize – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationSamplingParameters& the reference of this object.

class UnsuperSegmentationModelParameters : public visionflow::param::ISchemable

UnsuperSegmentationModelParameters Parameter class generated by jinja2 automatically.

Unsuper Segmentation Model Parameters Group.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const std::string &get_model_arch() const

Get Model Architecture.

General Model training slow but get better result, using for complex image; Training Fast Model training fast but might get worse performance than General Model

See also

set_model_arch()

Returns:

const std::string & Model Architecture

UnsuperSegmentationModelParameters &set_model_arch(const std::string &model_arch)

Set Model Architecture with std::string value.

General Model training slow but get better result, using for complex image; Training Fast Model training fast but might get worse performance than General Model

See also

get_model_arch()

Parameters:

model_arch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationModelParameters& the reference of this object.

bool get_add_logical_model() const

Get Logical Defect Detect.

increase detect performance for logical defect,such as character lost or duplicate, but increasing training and inference speed

See also

set_add_logical_model()

Returns:

bool Logical Defect Detect

UnsuperSegmentationModelParameters &set_add_logical_model(bool add_logical_model)

Set Logical Defect Detect with bool value.

increase detect performance for logical defect,such as character lost or duplicate, but increasing training and inference speed

See also

get_add_logical_model()

Parameters:

add_logical_model – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationModelParameters& the reference of this object.

Unsuper Classification Training Parameters

class UnsuperClassificationInputShape : public visionflow::param::ISchemable

UnsuperClassificationInputShape Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Custom Input Shape.

Enable to Customize Network Input Shape

See also

set_enable()

Returns:

bool Enable Custom Input Shape

UnsuperClassificationInputShape &set_enable(bool enable)

Set Enable Custom Input Shape with bool value.

Enable to Customize Network Input Shape

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperClassificationInputShape& the reference of this object.

int get_base_input_width() const

Get Base Input Width.

Base Input Width

See also

set_base_input_width()

Returns:

int Base Input Width

UnsuperClassificationInputShape &set_base_input_width(int base_input_width)

Set Base Input Width with int value.

Base Input Width

See also

get_base_input_width()

Parameters:

base_input_width – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperClassificationInputShape& the reference of this object.

int get_base_input_height() const

Get Base Input Height.

Base Input Height

See also

set_base_input_height()

Returns:

int Base Input Height

UnsuperClassificationInputShape &set_base_input_height(int base_input_height)

Set Base Input Height with int value.

Base Input Height

See also

get_base_input_height()

Parameters:

base_input_height – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperClassificationInputShape& the reference of this object.

class UnsuperClassificationTrainingParameters : public visionflow::param::SchemableParameter

UnsuperClassificationTrainingParameters Parameter class generated by jinja2 automatically.

Unsuper Classification Training Parameters Group.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int get_epoch() const

Get Epoch.

Number of epochs to train

See also

set_epoch()

Returns:

int Epoch

UnsuperClassificationTrainingParameters &set_epoch(int epoch)

Set Epoch with int value.

Number of epochs to train

See also

get_epoch()

Parameters:

epoch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperClassificationTrainingParameters& the reference of this object.

int get_batch_size() const

Get Training Batch Size.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16. The classification module needs to be set larger, generally 32, 64

See also

set_batch_size()

Returns:

int Training Batch Size

UnsuperClassificationTrainingParameters &set_batch_size(int batch_size)

Set Training Batch Size with int value.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16. The classification module needs to be set larger, generally 32, 64

See also

get_batch_size()

Parameters:

batch_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperClassificationTrainingParameters& the reference of this object.

const UnsuperClassificationInputShape &get_input_shape() const

Get Input Shape.

Input Shape Parameters

See also

set_input_shape()

Returns:

const UnsuperClassificationInputShape & Input Shape

UnsuperClassificationTrainingParameters &set_input_shape(const UnsuperClassificationInputShape &input_shape)

Set Input Shape with UnsuperClassificationInputShape value.

Input Shape Parameters

See also

get_input_shape()

Parameters:

input_shape – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperClassificationTrainingParameters& the reference of this object.

UnsuperClassificationInputShape &get_input_shape()

Get mutable reference of Input Shape.

Input Shape Parameters

Returns:

UnsuperClassificationTrainingParameters& the mutable reference of the group.

const std::string &get_model_arch() const

Get Model Architecture.

Currently only one baseline model

See also

set_model_arch()

Returns:

const std::string & Model Architecture

UnsuperClassificationTrainingParameters &set_model_arch(const std::string &model_arch)

Set Model Architecture with std::string value.

Currently only one baseline model

See also

get_model_arch()

Parameters:

model_arch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperClassificationTrainingParameters& the reference of this object.

const DataAugmentation &get_augmentations() const

Get Data Augmentation.

Data Augmentation Parameters

See also

set_augmentations()

Returns:

const DataAugmentation & Data Augmentation

UnsuperClassificationTrainingParameters &set_augmentations(const DataAugmentation &augmentations)

Set Data Augmentation with DataAugmentation value.

Data Augmentation Parameters

See also

get_augmentations()

Parameters:

augmentations – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperClassificationTrainingParameters& the reference of this object.

DataAugmentation &get_augmentations()

Get mutable reference of Data Augmentation.

Data Augmentation Parameters

Returns:

UnsuperClassificationTrainingParameters& the mutable reference of the group.

Unsuper Segmentation Inference Parameters

class UnsuperDefectRadius : public visionflow::param::ISchemable

UnsuperDefectRadius Parameter class generated by jinja2 automatically.

Parameters using for define defect size

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_auto_set() const

Get Automatic Setting.

Network adjust defect radius automatically

See also

set_auto_set()

Returns:

bool Automatic Setting

UnsuperDefectRadius &set_auto_set(bool auto_set)

Set Automatic Setting with bool value.

Network adjust defect radius automatically

See also

get_auto_set()

Parameters:

auto_set – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperDefectRadius& the reference of this object.

int get_size() const

Get Defect Radius.

defect size

See also

set_size()

Returns:

int Defect Radius

UnsuperDefectRadius &set_size(int size)

Set Defect Radius with int value.

defect size

See also

get_size()

Parameters:

size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperDefectRadius& the reference of this object.

class UnsuperSegmentationInferenceParameters : public visionflow::param::SchemableParameter

UnsuperSegmentationInferenceParameters Parameter class generated by jinja2 automatically.

UnsuperSegmentationInferenceParameters

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int get_minimum_defect_size() const

Get Minimum Defect Size.

Minimum detect that customer want to detect, but too small will increase overkill

See also

set_minimum_defect_size()

Returns:

int Minimum Defect Size

UnsuperSegmentationInferenceParameters &set_minimum_defect_size(int minimum_defect_size)

Set Minimum Defect Size with int value.

Minimum detect that customer want to detect, but too small will increase overkill

See also

get_minimum_defect_size()

Parameters:

minimum_defect_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationInferenceParameters& the reference of this object.

int get_sampling_density() const

Get Sampling Density.

smaller value will make result more precise, but increase inference time

See also

set_sampling_density()

Returns:

int Sampling Density

UnsuperSegmentationInferenceParameters &set_sampling_density(int sampling_density)

Set Sampling Density with int value.

smaller value will make result more precise, but increase inference time

See also

get_sampling_density()

Parameters:

sampling_density – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperSegmentationInferenceParameters& the reference of this object.

Unsuper Classification Inference Parameters

class UnsuperClassificationInferenceParameters : public visionflow::param::SchemableParameter

UnsuperClassificationInferenceParameters Parameter class generated by jinja2 automatically.

Parameters using for unsuper classification inference

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

double get_ng_thresh() const

Get Ng Thresh.

image will be set to ok when score < Ng Thresh, and set to ng when score > Ng Thresh

See also

set_ng_thresh()

Returns:

double Ng Thresh

UnsuperClassificationInferenceParameters &set_ng_thresh(double ng_thresh)

Set Ng Thresh with double value.

image will be set to ok when score < Ng Thresh, and set to ng when score > Ng Thresh

See also

get_ng_thresh()

Parameters:

ng_thresh – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperClassificationInferenceParameters& the reference of this object.

const UnsuperDefectRadius &get_defect_radius() const

Get Defect Radius.

Expected defect size

See also

set_defect_radius()

Returns:

const UnsuperDefectRadius & Defect Radius

UnsuperClassificationInferenceParameters &set_defect_radius(const UnsuperDefectRadius &defect_radius)

Set Defect Radius with UnsuperDefectRadius value.

Expected defect size

See also

get_defect_radius()

Parameters:

defect_radius – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

UnsuperClassificationInferenceParameters& the reference of this object.

UnsuperDefectRadius &get_defect_radius()

Get mutable reference of Defect Radius.

Expected defect size

Returns:

UnsuperClassificationInferenceParameters& the mutable reference of the group.

Filter Parameters

class AxialSideLengthRange : public visionflow::param::ISchemable

AxialSideLengthRange Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Axial Side Length Filter.

Enable this option to filter regions with axial side length.

See also

set_enable()

Returns:

bool Axial Side Length Filter

AxialSideLengthRange &set_enable(bool enable)

Set Axial Side Length Filter with bool value.

Enable this option to filter regions with axial side length.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AxialSideLengthRange& the reference of this object.

const std::vector<int> &get_x_axis_side_range() const

Get X-Axis Side Length Range.

The X-axis side length range of regions you want to keep.

See also

set_x_axis_side_range()

Returns:

const std::vector<int> & X-Axis Side Length Range

AxialSideLengthRange &set_x_axis_side_range(const std::vector<int> &x_axis_side_range)

Set X-Axis Side Length Range with std::vector<int> value.

The X-axis side length range of regions you want to keep.

See also

get_x_axis_side_range()

Parameters:

x_axis_side_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AxialSideLengthRange& the reference of this object.

bool x_axis_side_range_contains(int value) const

Check if the X-Axis Side Length Range contains the value.

Returns:

bool true if the X-Axis Side Length Range contains the value, otherwise false.

int get_x_axis_side_range_left() const

Get left point value of X-Axis Side Length Range.

See also

set_x_axis_side_range_right()

Returns:

const std::vector<int> & left point value of X-Axis Side Length Range

int get_x_axis_side_range_right() const

Get the right point value of X-Axis Side Length Range.

See also

set_x_axis_side_range_left()

Returns:

const std::vector<int> & the right point value of X-Axis Side Length Range

AxialSideLengthRange &set_x_axis_side_range_left(int x_axis_side_range_left)

Set left point value of X-Axis Side Length Range with std::vector<int> value.

The X-axis side length range of regions you want to keep.

See also

get_x_axis_side_range_left()

Parameters:

x_axis_side_range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AxialSideLengthRange& the reference of this object.

AxialSideLengthRange &set_x_axis_side_range_right(int x_axis_side_range_right)

Set the right point value of X-Axis Side Length Range with std::vector<int> value.

The X-axis side length range of regions you want to keep.

See also

get_x_axis_side_range_right()

Parameters:

x_axis_side_range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AxialSideLengthRange& the reference of this object.

const std::vector<int> &get_y_axis_side_range() const

Get Y-Axis Side Length Range.

The Y-axis side length range of regions you want to keep.

See also

set_y_axis_side_range()

Returns:

const std::vector<int> & Y-Axis Side Length Range

AxialSideLengthRange &set_y_axis_side_range(const std::vector<int> &y_axis_side_range)

Set Y-Axis Side Length Range with std::vector<int> value.

The Y-axis side length range of regions you want to keep.

See also

get_y_axis_side_range()

Parameters:

y_axis_side_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AxialSideLengthRange& the reference of this object.

bool y_axis_side_range_contains(int value) const

Check if the Y-Axis Side Length Range contains the value.

Returns:

bool true if the Y-Axis Side Length Range contains the value, otherwise false.

int get_y_axis_side_range_left() const

Get left point value of Y-Axis Side Length Range.

See also

set_y_axis_side_range_right()

Returns:

const std::vector<int> & left point value of Y-Axis Side Length Range

int get_y_axis_side_range_right() const

Get the right point value of Y-Axis Side Length Range.

See also

set_y_axis_side_range_left()

Returns:

const std::vector<int> & the right point value of Y-Axis Side Length Range

AxialSideLengthRange &set_y_axis_side_range_left(int y_axis_side_range_left)

Set left point value of Y-Axis Side Length Range with std::vector<int> value.

The Y-axis side length range of regions you want to keep.

See also

get_y_axis_side_range_left()

Parameters:

y_axis_side_range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AxialSideLengthRange& the reference of this object.

AxialSideLengthRange &set_y_axis_side_range_right(int y_axis_side_range_right)

Set the right point value of Y-Axis Side Length Range with std::vector<int> value.

The Y-axis side length range of regions you want to keep.

See also

get_y_axis_side_range_right()

Parameters:

y_axis_side_range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AxialSideLengthRange& the reference of this object.

const std::vector<double> &get_x_y_ratio_range() const

Get X/Y Ratio Range.

The range of the ratio of X-axis side length to the Y-axis side length of the regions that you want to keep.

See also

set_x_y_ratio_range()

Returns:

const std::vector<double> & X/Y Ratio Range

AxialSideLengthRange &set_x_y_ratio_range(const std::vector<double> &x_y_ratio_range)

Set X/Y Ratio Range with std::vector<double> value.

The range of the ratio of X-axis side length to the Y-axis side length of the regions that you want to keep.

See also

get_x_y_ratio_range()

Parameters:

x_y_ratio_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AxialSideLengthRange& the reference of this object.

bool x_y_ratio_range_contains(double value) const

Check if the X/Y Ratio Range contains the value.

Returns:

bool true if the X/Y Ratio Range contains the value, otherwise false.

double get_x_y_ratio_range_left() const

Get left point value of X/Y Ratio Range.

See also

set_x_y_ratio_range_right()

Returns:

const std::vector<double> & left point value of X/Y Ratio Range

double get_x_y_ratio_range_right() const

Get the right point value of X/Y Ratio Range.

See also

set_x_y_ratio_range_left()

Returns:

const std::vector<double> & the right point value of X/Y Ratio Range

AxialSideLengthRange &set_x_y_ratio_range_left(double x_y_ratio_range_left)

Set left point value of X/Y Ratio Range with std::vector<double> value.

The range of the ratio of X-axis side length to the Y-axis side length of the regions that you want to keep.

See also

get_x_y_ratio_range_left()

Parameters:

x_y_ratio_range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AxialSideLengthRange& the reference of this object.

AxialSideLengthRange &set_x_y_ratio_range_right(double x_y_ratio_range_right)

Set the right point value of X/Y Ratio Range with std::vector<double> value.

The range of the ratio of X-axis side length to the Y-axis side length of the regions that you want to keep.

See also

get_x_y_ratio_range_right()

Parameters:

x_y_ratio_range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AxialSideLengthRange& the reference of this object.

class SideLengthRange : public visionflow::param::ISchemable

SideLengthRange Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Side Length Filter.

Enable this option to filter regions with long side length and short side length.

See also

set_enable()

Returns:

bool Side Length Filter

SideLengthRange &set_enable(bool enable)

Set Side Length Filter with bool value.

Enable this option to filter regions with long side length and short side length.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SideLengthRange& the reference of this object.

bool get_on_rotate_rect() const

Get Filter On Min-Area-Rect.

The “Longer Side Length Range” and “Shorter Side Length Range” will be applied to the min-area rotate rectangle of each region if this option is enabled, otherwise, they will be applied to the axial bounding rect.

See also

set_on_rotate_rect()

Returns:

bool Filter On Min-Area-Rect

SideLengthRange &set_on_rotate_rect(bool on_rotate_rect)

Set Filter On Min-Area-Rect with bool value.

The “Longer Side Length Range” and “Shorter Side Length Range” will be applied to the min-area rotate rectangle of each region if this option is enabled, otherwise, they will be applied to the axial bounding rect.

See also

get_on_rotate_rect()

Parameters:

on_rotate_rect – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SideLengthRange& the reference of this object.

const std::vector<int> &get_longer_side_range() const

Get Longer Side Length Range.

The longer side length range of regions you want to keep.

See also

set_longer_side_range()

Returns:

const std::vector<int> & Longer Side Length Range

SideLengthRange &set_longer_side_range(const std::vector<int> &longer_side_range)

Set Longer Side Length Range with std::vector<int> value.

The longer side length range of regions you want to keep.

See also

get_longer_side_range()

Parameters:

longer_side_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SideLengthRange& the reference of this object.

bool longer_side_range_contains(int value) const

Check if the Longer Side Length Range contains the value.

Returns:

bool true if the Longer Side Length Range contains the value, otherwise false.

int get_longer_side_range_left() const

Get left point value of Longer Side Length Range.

See also

set_longer_side_range_right()

Returns:

const std::vector<int> & left point value of Longer Side Length Range

int get_longer_side_range_right() const

Get the right point value of Longer Side Length Range.

See also

set_longer_side_range_left()

Returns:

const std::vector<int> & the right point value of Longer Side Length Range

SideLengthRange &set_longer_side_range_left(int longer_side_range_left)

Set left point value of Longer Side Length Range with std::vector<int> value.

The longer side length range of regions you want to keep.

See also

get_longer_side_range_left()

Parameters:

longer_side_range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SideLengthRange& the reference of this object.

SideLengthRange &set_longer_side_range_right(int longer_side_range_right)

Set the right point value of Longer Side Length Range with std::vector<int> value.

The longer side length range of regions you want to keep.

See also

get_longer_side_range_right()

Parameters:

longer_side_range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SideLengthRange& the reference of this object.

const std::vector<int> &get_shorter_side_range() const

Get Shorter Side Length Range.

The shorter side length range of regions you want to keep.

See also

set_shorter_side_range()

Returns:

const std::vector<int> & Shorter Side Length Range

SideLengthRange &set_shorter_side_range(const std::vector<int> &shorter_side_range)

Set Shorter Side Length Range with std::vector<int> value.

The shorter side length range of regions you want to keep.

See also

get_shorter_side_range()

Parameters:

shorter_side_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SideLengthRange& the reference of this object.

bool shorter_side_range_contains(int value) const

Check if the Shorter Side Length Range contains the value.

Returns:

bool true if the Shorter Side Length Range contains the value, otherwise false.

int get_shorter_side_range_left() const

Get left point value of Shorter Side Length Range.

See also

set_shorter_side_range_right()

Returns:

const std::vector<int> & left point value of Shorter Side Length Range

int get_shorter_side_range_right() const

Get the right point value of Shorter Side Length Range.

See also

set_shorter_side_range_left()

Returns:

const std::vector<int> & the right point value of Shorter Side Length Range

SideLengthRange &set_shorter_side_range_left(int shorter_side_range_left)

Set left point value of Shorter Side Length Range with std::vector<int> value.

The shorter side length range of regions you want to keep.

See also

get_shorter_side_range_left()

Parameters:

shorter_side_range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SideLengthRange& the reference of this object.

SideLengthRange &set_shorter_side_range_right(int shorter_side_range_right)

Set the right point value of Shorter Side Length Range with std::vector<int> value.

The shorter side length range of regions you want to keep.

See also

get_shorter_side_range_right()

Parameters:

shorter_side_range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SideLengthRange& the reference of this object.

const std::vector<double> &get_aspect_ratio_range() const

Get Aspect Ratio Range.

The range of the ratio of the short side to the long side of the regions that you want to keep.

See also

set_aspect_ratio_range()

Returns:

const std::vector<double> & Aspect Ratio Range

SideLengthRange &set_aspect_ratio_range(const std::vector<double> &aspect_ratio_range)

Set Aspect Ratio Range with std::vector<double> value.

The range of the ratio of the short side to the long side of the regions that you want to keep.

See also

get_aspect_ratio_range()

Parameters:

aspect_ratio_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SideLengthRange& the reference of this object.

bool aspect_ratio_range_contains(double value) const

Check if the Aspect Ratio Range contains the value.

Returns:

bool true if the Aspect Ratio Range contains the value, otherwise false.

double get_aspect_ratio_range_left() const

Get left point value of Aspect Ratio Range.

See also

set_aspect_ratio_range_right()

Returns:

const std::vector<double> & left point value of Aspect Ratio Range

double get_aspect_ratio_range_right() const

Get the right point value of Aspect Ratio Range.

See also

set_aspect_ratio_range_left()

Returns:

const std::vector<double> & the right point value of Aspect Ratio Range

SideLengthRange &set_aspect_ratio_range_left(double aspect_ratio_range_left)

Set left point value of Aspect Ratio Range with std::vector<double> value.

The range of the ratio of the short side to the long side of the regions that you want to keep.

See also

get_aspect_ratio_range_left()

Parameters:

aspect_ratio_range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SideLengthRange& the reference of this object.

SideLengthRange &set_aspect_ratio_range_right(double aspect_ratio_range_right)

Set the right point value of Aspect Ratio Range with std::vector<double> value.

The range of the ratio of the short side to the long side of the regions that you want to keep.

See also

get_aspect_ratio_range_right()

Parameters:

aspect_ratio_range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SideLengthRange& the reference of this object.

class FilterScript : public visionflow::param::ISchemable

FilterScript Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Filter function.

Enable this option to filter regions with customized python function.

See also

set_enable()

Returns:

bool Enable Filter function

FilterScript &set_enable(bool enable)

Set Enable Filter function with bool value.

Enable this option to filter regions with customized python function.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

FilterScript& the reference of this object.

const std::string &get_filter_script() const

Get Filter Script.

The python filter script.

See also

set_filter_script()

Returns:

const std::string & Filter Script

FilterScript &set_filter_script(const std::string &filter_script)

Set Filter Script with std::string value.

The python filter script.

See also

get_filter_script()

Parameters:

filter_script – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

FilterScript& the reference of this object.

class SingleClassPolygonsFilterParameters : public visionflow::param::ISchemable

SingleClassPolygonsFilterParameters Parameter class generated by jinja2 automatically.

Common Regions filter parameters to filter regions.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable.

Enable this polygon filter parameter group.

See also

set_enable()

Returns:

bool Enable

SingleClassPolygonsFilterParameters &set_enable(bool enable)

Set Enable with bool value.

Enable this polygon filter parameter group.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SingleClassPolygonsFilterParameters& the reference of this object.

double get_score_threshold() const

Get Score Threshold.

Only retains regions with score greater than this threshold.

See also

set_score_threshold()

Returns:

double Score Threshold

SingleClassPolygonsFilterParameters &set_score_threshold(double score_threshold)

Set Score Threshold with double value.

Only retains regions with score greater than this threshold.

See also

get_score_threshold()

Parameters:

score_threshold – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SingleClassPolygonsFilterParameters& the reference of this object.

const std::vector<int> &get_area_range() const

Get Area Range.

The area range of the regions you want to keep

See also

set_area_range()

Returns:

const std::vector<int> & Area Range

SingleClassPolygonsFilterParameters &set_area_range(const std::vector<int> &area_range)

Set Area Range with std::vector<int> value.

The area range of the regions you want to keep

See also

get_area_range()

Parameters:

area_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SingleClassPolygonsFilterParameters& the reference of this object.

bool area_range_contains(int value) const

Check if the Area Range contains the value.

Returns:

bool true if the Area Range contains the value, otherwise false.

int get_area_range_left() const

Get left point value of Area Range.

See also

set_area_range_right()

Returns:

const std::vector<int> & left point value of Area Range

int get_area_range_right() const

Get the right point value of Area Range.

See also

set_area_range_left()

Returns:

const std::vector<int> & the right point value of Area Range

SingleClassPolygonsFilterParameters &set_area_range_left(int area_range_left)

Set left point value of Area Range with std::vector<int> value.

The area range of the regions you want to keep

See also

get_area_range_left()

Parameters:

area_range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SingleClassPolygonsFilterParameters& the reference of this object.

SingleClassPolygonsFilterParameters &set_area_range_right(int area_range_right)

Set the right point value of Area Range with std::vector<int> value.

The area range of the regions you want to keep

See also

get_area_range_right()

Parameters:

area_range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SingleClassPolygonsFilterParameters& the reference of this object.

const AxialSideLengthRange &get_axial_side_filter() const

Get Axial-Side-Length Filter.

Filter regions with X/Y-axis side length.

See also

set_axial_side_filter()

Returns:

const AxialSideLengthRange & Axial-Side-Length Filter

SingleClassPolygonsFilterParameters &set_axial_side_filter(const AxialSideLengthRange &axial_side_filter)

Set Axial-Side-Length Filter with AxialSideLengthRange value.

Filter regions with X/Y-axis side length.

See also

get_axial_side_filter()

Parameters:

axial_side_filter – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SingleClassPolygonsFilterParameters& the reference of this object.

AxialSideLengthRange &get_axial_side_filter()

Get mutable reference of Axial-Side-Length Filter.

Filter regions with X/Y-axis side length.

Returns:

SingleClassPolygonsFilterParameters& the mutable reference of the group.

const SideLengthRange &get_side_filter() const

Get Side-Length Filter.

Filter regions with Side length

See also

set_side_filter()

Returns:

const SideLengthRange & Side-Length Filter

SingleClassPolygonsFilterParameters &set_side_filter(const SideLengthRange &side_filter)

Set Side-Length Filter with SideLengthRange value.

Filter regions with Side length

See also

get_side_filter()

Parameters:

side_filter – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SingleClassPolygonsFilterParameters& the reference of this object.

SideLengthRange &get_side_filter()

Get mutable reference of Side-Length Filter.

Filter regions with Side length

Returns:

SingleClassPolygonsFilterParameters& the mutable reference of the group.

class PolygonsFilterParameters : public visionflow::param::SchemableParameter

PolygonsFilterParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const std::map<std::string, SingleClassPolygonsFilterParameters> &get_class_thresholds() const

Get Class Polygon Thresholds.

Polygon thresholds (area, side-length,etc.) named with the label classes according to which to filter the polygons.

See also

set_class_thresholds()

Returns:

const std::map<std::string, SingleClassPolygonsFilterParameters> & Class Polygon Thresholds

PolygonsFilterParameters &set_class_thresholds(const std::map<std::string, SingleClassPolygonsFilterParameters> &class_thresholds)

Set Class Polygon Thresholds with std::map<std::string, SingleClassPolygonsFilterParameters> value.

Polygon thresholds (area, side-length,etc.) named with the label classes according to which to filter the polygons.

See also

get_class_thresholds()

Parameters:

class_thresholds – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

PolygonsFilterParameters& the reference of this object.

const SingleClassPolygonsFilterParameters &get_class_thresholds(const std::string &key) const

Get value in Class Polygon Thresholds with key.

Warning

The key must be exist in Class Polygon Thresholds. otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const std::map<std::string, SingleClassPolygonsFilterParameters> & value in Class Polygon Thresholds at key.

PolygonsFilterParameters &set_class_thresholds(const std::string &key, const SingleClassPolygonsFilterParameters &value)

Set value in Class Polygon Thresholds with key.

Parameters:

• key – the key to set.

• value – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

PolygonsFilterParameters& the reference of this object.

bool class_thresholds_contains(const std::string &key) const

Check if the key is exist in Class Polygon Thresholds.

Returns:

bool true if the key is exist in Class Polygon Thresholds, otherwise false.

size_t get_class_thresholds_size() const

Get the size of Class Polygon Thresholds.

Returns:

size_t the size of Class Polygon Thresholds

SingleClassPolygonsFilterParameters &get_class_thresholds(const std::string &key)

Get mutable reference of value in Class Polygon Thresholds with key. A new key and default value will be created if the key does not exist.

Returns:

SingleClassPolygonsFilterParameters& the mutable reference of value in Class Polygon Thresholds at key.

const SingleClassPolygonsFilterParameters &get_additional_threshold() const

Get Additional Probability Thresholds.

Additional polygon threshold for other label classes not in “Class Polygon Thresholds”. The filter operator will filter all feature maps whose label class is not in “Class Polygon Thresholds” with the additional threshold.

See also

set_additional_threshold()

Returns:

const SingleClassPolygonsFilterParameters & Additional Probability Thresholds

PolygonsFilterParameters &set_additional_threshold(const SingleClassPolygonsFilterParameters &additional_threshold)

Set Additional Probability Thresholds with SingleClassPolygonsFilterParameters value.

Additional polygon threshold for other label classes not in “Class Polygon Thresholds”. The filter operator will filter all feature maps whose label class is not in “Class Polygon Thresholds” with the additional threshold.

See also

get_additional_threshold()

Parameters:

additional_threshold – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

PolygonsFilterParameters& the reference of this object.

SingleClassPolygonsFilterParameters &get_additional_threshold()

Get mutable reference of Additional Probability Thresholds.

Additional polygon threshold for other label classes not in “Class Polygon Thresholds”. The filter operator will filter all feature maps whose label class is not in “Class Polygon Thresholds” with the additional threshold.

Returns:

PolygonsFilterParameters& the mutable reference of the group.

const FilterScript &get_script_filter() const

Get Custom python filter script.

Filter regions with python script.

See also

set_script_filter()

Returns:

const FilterScript & Custom python filter script

PolygonsFilterParameters &set_script_filter(const FilterScript &script_filter)

Set Custom python filter script with FilterScript value.

Filter regions with python script.

See also

get_script_filter()

Parameters:

script_filter – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

PolygonsFilterParameters& the reference of this object.

FilterScript &get_script_filter()

Get mutable reference of Custom python filter script.

Filter regions with python script.

Returns:

PolygonsFilterParameters& the mutable reference of the group.

class FeatureMapFilterParameters : public visionflow::param::SchemableParameter

FeatureMapFilterParameters Parameter class generated by jinja2 automatically.

Parameters to config the feature map filter.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const std::map<std::string, double> &get_class_thresholds() const

Get Class Probability Thresholds.

Thresholds named with the label classes according to which to binarize the feature map.

See also

set_class_thresholds()

Returns:

const std::map<std::string, double> & Class Probability Thresholds

FeatureMapFilterParameters &set_class_thresholds(const std::map<std::string, double> &class_thresholds)

Set Class Probability Thresholds with std::map<std::string, double> value.

Thresholds named with the label classes according to which to binarize the feature map.

See also

get_class_thresholds()

Parameters:

class_thresholds – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

FeatureMapFilterParameters& the reference of this object.

double get_class_thresholds(const std::string &key) const

Get value in Class Probability Thresholds with key.

Warning

The key must be exist in Class Probability Thresholds. otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const std::map<std::string, double> & value in Class Probability Thresholds at key.

FeatureMapFilterParameters &set_class_thresholds(const std::string &key, double value)

Set value in Class Probability Thresholds with key.

Parameters:

• key – the key to set.

• value – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

FeatureMapFilterParameters& the reference of this object.

bool class_thresholds_contains(const std::string &key) const

Check if the key is exist in Class Probability Thresholds.

Returns:

bool true if the key is exist in Class Probability Thresholds, otherwise false.

size_t get_class_thresholds_size() const

Get the size of Class Probability Thresholds.

Returns:

size_t the size of Class Probability Thresholds

double get_additional_threshold() const

Get Additional Probability Thresholds.

Additional threshold for other label classes not in “Class Probability Thresholds”. The visionflow::opers::FeatureMapFilter will filter all feature maps whose label class is not in “Class Probability Thresholds” with the additional threshold.

See also

set_additional_threshold()

Returns:

double Additional Probability Thresholds

FeatureMapFilterParameters &set_additional_threshold(double additional_threshold)

Set Additional Probability Thresholds with double value.

Additional threshold for other label classes not in “Class Probability Thresholds”. The visionflow::opers::FeatureMapFilter will filter all feature maps whose label class is not in “Class Probability Thresholds” with the additional threshold.

See also

get_additional_threshold()

Parameters:

additional_threshold – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

FeatureMapFilterParameters& the reference of this object.

int get_dilate_pixels() const

Get Dilate Pixels.

Dilate the filtered regions by the specified pixels. This option can be used to merge closely scattered defects together.

See also

set_dilate_pixels()

Returns:

int Dilate Pixels

FeatureMapFilterParameters &set_dilate_pixels(int dilate_pixels)

Set Dilate Pixels with int value.

Dilate the filtered regions by the specified pixels. This option can be used to merge closely scattered defects together.

See also

get_dilate_pixels()

Parameters:

dilate_pixels – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

FeatureMapFilterParameters& the reference of this object.

bool get_enable_polygon_score() const

Get Calculate Defect Region Score.

Provide every defect region score(0~1), bigger value means the region has more possibility to be a defect

See also

set_enable_polygon_score()

Returns:

bool Calculate Defect Region Score

FeatureMapFilterParameters &set_enable_polygon_score(bool enable_polygon_score)

Set Calculate Defect Region Score with bool value.

Provide every defect region score(0~1), bigger value means the region has more possibility to be a defect

See also

get_enable_polygon_score()

Parameters:

enable_polygon_score – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

FeatureMapFilterParameters& the reference of this object.

Training logs

class LossCurve : public visionflow::param::ISchemable

LossCurve Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const std::vector<double> &get_values() const

Get values.

A loss curve

See also

set_values()

Returns:

const std::vector<double> & values

LossCurve &set_values(const std::vector<double> &values)

Set values with std::vector<double> value.

A loss curve

See also

get_values()

Parameters:

values – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LossCurve& the reference of this object.

double get_values(size_t index) const

Get value in values with index.

Warning

The index must be less than get_values_size(). otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

double value in values at index.

size_t get_values_size() const

Get the size of values.

Returns:

size_t the size of values

class TrainingLog : public visionflow::param::SchemableParameter

TrainingLog Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int64_t get_start_at() const

Get Start Time.

The training task start time.

See also

set_start_at()

Returns:

int64_t Start Time

TrainingLog &set_start_at(int64_t start_at)

Set Start Time with int64_t value.

The training task start time.

See also

get_start_at()

Parameters:

start_at – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

TrainingLog& the reference of this object.

int64_t get_end_at() const

Get End Time.

The training task finish time.

See also

set_end_at()

Returns:

int64_t End Time

TrainingLog &set_end_at(int64_t end_at)

Set End Time with int64_t value.

The training task finish time.

See also

get_end_at()

Parameters:

end_at – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

TrainingLog& the reference of this object.

const std::map<std::string, LossCurve> &get_loss() const

Get Loss.

Loss curves with name

See also

set_loss()

Returns:

const std::map<std::string, LossCurve> & Loss

TrainingLog &set_loss(const std::map<std::string, LossCurve> &loss)

Set Loss with std::map<std::string, LossCurve> value.

Loss curves with name

See also

get_loss()

Parameters:

loss – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

TrainingLog& the reference of this object.

const LossCurve &get_loss(const std::string &key) const

Get value in Loss with key.

Warning

The key must be exist in Loss. otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const std::map<std::string, LossCurve> & value in Loss at key.

TrainingLog &set_loss(const std::string &key, const LossCurve &value)

Set value in Loss with key.

Parameters:

• key – the key to set.

• value – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

TrainingLog& the reference of this object.

bool loss_contains(const std::string &key) const

Check if the key is exist in Loss.

Returns:

bool true if the key is exist in Loss, otherwise false.

size_t get_loss_size() const

Get the size of Loss.

Returns:

size_t the size of Loss

LossCurve &get_loss(const std::string &key)

Get mutable reference of value in Loss with key. A new key and default value will be created if the key does not exist.

Returns:

LossCurve& the mutable reference of value in Loss at key.

Training Mode

enum visionflow::param::TrainingMode

Values:

enumerator kNormalTrain = 0

enumerator kIncrementalTrain = 1

Ungrouped Parameters

class InferenceBatchSize : public visionflow::param::SchemableParameter

InferenceBatchSize Parameter class generated by jinja2 automatically.

Inference BatchSize, Currently only contains batch size. It may need to be refactored in the future.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int get_batch_size() const

Get Inference Batch Size.

The batch size for inference. You can set it to the number of pictures you wish to infer at the same time. Usually the default value of 1 is sufficient. Note that if you set larger batches without giving enough images per inference, it may actually slow down inference.

See also

set_batch_size()

Returns:

int Inference Batch Size

InferenceBatchSize &set_batch_size(int batch_size)

Set Inference Batch Size with int value.

The batch size for inference. You can set it to the number of pictures you wish to infer at the same time. Usually the default value of 1 is sufficient. Note that if you set larger batches without giving enough images per inference, it may actually slow down inference.

See also

get_batch_size()

Parameters:

batch_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

InferenceBatchSize& the reference of this object.

class TRTCalibParameters : public visionflow::param::SchemableParameter

TRTCalibParameters Parameter class generated by jinja2 automatically.

TensorRT Int8 Calibrator Parameters

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_place_holder() const

Get PlaceHolder.

See also

set_place_holder()

Returns:

bool PlaceHolder

TRTCalibParameters &set_place_holder(bool place_holder)

Set PlaceHolder with bool value.

See also

get_place_holder()

Parameters:

place_holder – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

TRTCalibParameters& the reference of this object.

Template Match Parameters

enum visionflow::param::TemplateMode

Values:

enumerator kValueByPixel = 0

enumerator kValueByRatio = 1

class KeyPointNode : public visionflow::param::ISchemable

KeyPointNode Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

double get_x() const

Get X-Value.

X-direction coordinate value of the center of the key point. When the template models set to kValueByPixel, X-Value should be set using pixel value. When the template models set to kValueByRatio, X-Value should be set using ratio value. When the Area Value Type is modified, X-Value needs to be modified.

See also

set_x()

Returns:

double X-Value

KeyPointNode &set_x(double x)

Set X-Value with double value.

X-direction coordinate value of the center of the key point. When the template models set to kValueByPixel, X-Value should be set using pixel value. When the template models set to kValueByRatio, X-Value should be set using ratio value. When the Area Value Type is modified, X-Value needs to be modified.

See also

get_x()

Parameters:

x – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

KeyPointNode& the reference of this object.

double get_y() const

Get Y-Value.

Y-direction coordinate value of the center of the key point. When the template models set to kValueByPixel, Y-Value should be set using pixel value. When the template models set to kValueByRatio, Y-Value should be set using ratio value. When the Area Value Type is modified, Y-Value needs to be modified.

See also

set_y()

Returns:

double Y-Value

KeyPointNode &set_y(double y)

Set Y-Value with double value.

Y-direction coordinate value of the center of the key point. When the template models set to kValueByPixel, Y-Value should be set using pixel value. When the template models set to kValueByRatio, Y-Value should be set using ratio value. When the Area Value Type is modified, Y-Value needs to be modified.

See also

get_y()

Parameters:

y – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

KeyPointNode& the reference of this object.

const std::string &get_name() const

Get Name.

The name that can be matched by the node

See also

set_name()

Returns:

const std::string & Name

KeyPointNode &set_name(const std::string &name)

Set Name with std::string value.

The name that can be matched by the node

See also

get_name()

Parameters:

name – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

KeyPointNode& the reference of this object.

class MultiNameKeyPointNode : public visionflow::param::ISchemable

MultiNameKeyPointNode Parameter class generated by jinja2 automatically.

Define the position of the node and a set of names that can be matched by the node.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

double get_x() const

Get X-Value.

X-direction coordinate value of the key point, specified either in absolute pixel or relative ratio, depending on the chosen TemplateMode. If using kValueByRatio mode, the value is relative to the view width.

See also

set_x()

Returns:

double X-Value

MultiNameKeyPointNode &set_x(double x)

Set X-Value with double value.

X-direction coordinate value of the key point, specified either in absolute pixel or relative ratio, depending on the chosen TemplateMode. If using kValueByRatio mode, the value is relative to the view width.

See also

get_x()

Parameters:

x – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

MultiNameKeyPointNode& the reference of this object.

double get_y() const

Get Y-Value.

Y-direction coordinate value of the key point, specified either in absolute pixel or relative ratio, depending on the chosen TemplateMode. If using kValueByRatio mode, the value is relative to the view height.

See also

set_y()

Returns:

double Y-Value

MultiNameKeyPointNode &set_y(double y)

Set Y-Value with double value.

Y-direction coordinate value of the key point, specified either in absolute pixel or relative ratio, depending on the chosen TemplateMode. If using kValueByRatio mode, the value is relative to the view height.

See also

get_y()

Parameters:

y – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

MultiNameKeyPointNode& the reference of this object.

const std::vector<std::string> &get_names() const

Get Names.

The names that can be matched by the node

See also

set_names()

Returns:

const std::vector<std::string> & Names

MultiNameKeyPointNode &set_names(const std::vector<std::string> &names)

Set Names with std::vector<std::string> value.

The names that can be matched by the node

See also

get_names()

Parameters:

names – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

MultiNameKeyPointNode& the reference of this object.

const std::string &get_names(size_t index) const

Get value in Names with index.

Warning

The index must be less than get_names_size(). otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const std::string & value in Names at index.

size_t get_names_size() const

Get the size of Names.

Returns:

size_t the size of Names

OCR Parameters

class RegularExpression : public visionflow::param::ISchemable

RegularExpression Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Regular Expression.

Enable regular expression.

See also

set_enable()

Returns:

bool Enable Regular Expression

RegularExpression &set_enable(bool enable)

Set Enable Regular Expression with bool value.

Enable regular expression.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

RegularExpression& the reference of this object.

const std::string &get_pattern() const

Get Pattern.

Regular expression pattern.

See also

set_pattern()

Returns:

const std::string & Pattern

RegularExpression &set_pattern(const std::string &pattern)

Set Pattern with std::string value.

Regular expression pattern.

See also

get_pattern()

Parameters:

pattern – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

RegularExpression& the reference of this object.

class OCRNodeTemplate : public visionflow::param::ISchemable

OCRNodeTemplate Parameter class generated by jinja2 automatically.

Node template is used to match strings of any shape.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Node Template Matching.

Enable node template matching.

See also

set_enable()

Returns:

bool Enable Node Template Matching

OCRNodeTemplate &set_enable(bool enable)

Set Enable Node Template Matching with bool value.

Enable node template matching.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRNodeTemplate& the reference of this object.

const TemplateMode &get_template_mode() const

Get Template Mode.

Specifying the unit of measurement for the position and max_distance values of the key point node. Choose between kValueByPixel for using pixel values, or kValueByRatio for using relative ratios.

See also

set_template_mode()

Returns:

const TemplateMode & Template Mode

OCRNodeTemplate &set_template_mode(const TemplateMode &template_mode)

Set Template Mode with TemplateMode value.

Specifying the unit of measurement for the position and max_distance values of the key point node. Choose between kValueByPixel for using pixel values, or kValueByRatio for using relative ratios.

See also

get_template_mode()

Parameters:

template_mode – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRNodeTemplate& the reference of this object.

int get_max_matches() const

Get Maximum Matches.

Maximum number of templates to remain within view. Results are sorted based on the match distance in ascending order, and only the top ‘max_matches’ results are retained. All results are provided by default.

See also

set_max_matches()

Returns:

int Maximum Matches

OCRNodeTemplate &set_max_matches(int max_matches)

Set Maximum Matches with int value.

Maximum number of templates to remain within view. Results are sorted based on the match distance in ascending order, and only the top ‘max_matches’ results are retained. All results are provided by default.

See also

get_max_matches()

Parameters:

max_matches – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRNodeTemplate& the reference of this object.

const std::vector<MultiNameKeyPointNode> &get_points() const

Get Keypoints.

The key points of the template.

See also

set_points()

Returns:

const std::vector<MultiNameKeyPointNode> & Keypoints

OCRNodeTemplate &set_points(const std::vector<MultiNameKeyPointNode> &points)

Set Keypoints with std::vector<MultiNameKeyPointNode> value.

The key points of the template.

See also

get_points()

Parameters:

points – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRNodeTemplate& the reference of this object.

const MultiNameKeyPointNode &get_points(size_t index) const

Get value in Keypoints with index.

Warning

The index must be less than get_points_size(). otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const MultiNameKeyPointNode & value in Keypoints at index.

size_t get_points_size() const

Get the size of Keypoints.

Returns:

size_t the size of Keypoints

MultiNameKeyPointNode &get_points(size_t index)

Get mutable reference of value in Keypoints with index.

Warning

The index must be less than get_points_size(). otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

MultiNameKeyPointNode& the mutable reference of value in Keypoints at index.

double get_max_distance() const

Get Max Distance.

The maximum allowed distance between the template keypoints and the keypoints to be matched. Results with a distance greater than this value will be filtered out. The value can be either in absolute pixel or relative to the view height, depending on the template_mode.

See also

set_max_distance()

Returns:

double Max Distance

OCRNodeTemplate &set_max_distance(double max_distance)

Set Max Distance with double value.

The maximum allowed distance between the template keypoints and the keypoints to be matched. Results with a distance greater than this value will be filtered out. The value can be either in absolute pixel or relative to the view height, depending on the template_mode.

See also

get_max_distance()

Parameters:

max_distance – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRNodeTemplate& the reference of this object.

bool get_is_scalable() const

Get Target Scalable.

Enable this option if the size of target to be matched is not fixed. And you should set the target scale range.

See also

set_is_scalable()

Returns:

bool Target Scalable

OCRNodeTemplate &set_is_scalable(bool is_scalable)

Set Target Scalable with bool value.

Enable this option if the size of target to be matched is not fixed. And you should set the target scale range.

See also

get_is_scalable()

Parameters:

is_scalable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRNodeTemplate& the reference of this object.

const std::vector<double> &get_scale_range() const

Get Scale Range.

The target scale range.

See also

set_scale_range()

Returns:

const std::vector<double> & Scale Range

OCRNodeTemplate &set_scale_range(const std::vector<double> &scale_range)

Set Scale Range with std::vector<double> value.

The target scale range.

See also

get_scale_range()

Parameters:

scale_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRNodeTemplate& the reference of this object.

bool scale_range_contains(double value) const

Check if the Scale Range contains the value.

Returns:

bool true if the Scale Range contains the value, otherwise false.

double get_scale_range_left() const

Get left point value of Scale Range.

See also

set_scale_range_right()

Returns:

const std::vector<double> & left point value of Scale Range

double get_scale_range_right() const

Get the right point value of Scale Range.

See also

set_scale_range_left()

Returns:

const std::vector<double> & the right point value of Scale Range

OCRNodeTemplate &set_scale_range_left(double scale_range_left)

Set left point value of Scale Range with std::vector<double> value.

The target scale range.

See also

get_scale_range_left()

Parameters:

scale_range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRNodeTemplate& the reference of this object.

OCRNodeTemplate &set_scale_range_right(double scale_range_right)

Set the right point value of Scale Range with std::vector<double> value.

The target scale range.

See also

get_scale_range_right()

Parameters:

scale_range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRNodeTemplate& the reference of this object.

bool get_is_rotatable() const

Get Target Rotatable.

Enable this option if the angle of target to be matched is not fixed, And you should set the target rotate angle range.

See also

set_is_rotatable()

Returns:

bool Target Rotatable

OCRNodeTemplate &set_is_rotatable(bool is_rotatable)

Set Target Rotatable with bool value.

Enable this option if the angle of target to be matched is not fixed, And you should set the target rotate angle range.

See also

get_is_rotatable()

Parameters:

is_rotatable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRNodeTemplate& the reference of this object.

const std::vector<int> &get_rotate_range() const

Get Rotate Range (Degree).

The target rotate range.

See also

set_rotate_range()

Returns:

const std::vector<int> & Rotate Range (Degree)

OCRNodeTemplate &set_rotate_range(const std::vector<int> &rotate_range)

Set Rotate Range (Degree) with std::vector<int> value.

The target rotate range.

See also

get_rotate_range()

Parameters:

rotate_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRNodeTemplate& the reference of this object.

bool rotate_range_contains(int value) const

Check if the Rotate Range (Degree) contains the value.

Returns:

bool true if the Rotate Range (Degree) contains the value, otherwise false.

int get_rotate_range_left() const

Get left point value of Rotate Range (Degree).

See also

set_rotate_range_right()

Returns:

const std::vector<int> & left point value of Rotate Range (Degree)

int get_rotate_range_right() const

Get the right point value of Rotate Range (Degree).

See also

set_rotate_range_left()

Returns:

const std::vector<int> & the right point value of Rotate Range (Degree)

OCRNodeTemplate &set_rotate_range_left(int rotate_range_left)

Set left point value of Rotate Range (Degree) with std::vector<int> value.

The target rotate range.

See also

get_rotate_range_left()

Parameters:

rotate_range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRNodeTemplate& the reference of this object.

OCRNodeTemplate &set_rotate_range_right(int rotate_range_right)

Set the right point value of Rotate Range (Degree) with std::vector<int> value.

The target rotate range.

See also

get_rotate_range_right()

Parameters:

rotate_range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRNodeTemplate& the reference of this object.

class OCRStringTemplate : public visionflow::param::ISchemable

OCRStringTemplate Parameter class generated by jinja2 automatically.

String template is used to control how to match characters into strings.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable String Template Matching.

Enable string template matching.

See also

set_enable()

Returns:

bool Enable String Template Matching

OCRStringTemplate &set_enable(bool enable)

Set Enable String Template Matching with bool value.

Enable string template matching.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRStringTemplate& the reference of this object.

int get_max_matches() const

Get Maximum Matches.

Maximum number of templates to remain within view. Results are sorted based on the deviation of characters from the string, and only the top ‘max_matches’ results are retained. All results are provided by default.

See also

set_max_matches()

Returns:

int Maximum Matches

OCRStringTemplate &set_max_matches(int max_matches)

Set Maximum Matches with int value.

Maximum number of templates to remain within view. Results are sorted based on the deviation of characters from the string, and only the top ‘max_matches’ results are retained. All results are provided by default.

See also

get_max_matches()

Parameters:

max_matches – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRStringTemplate& the reference of this object.

int get_min_length() const

Get Minimum String Length.

String shorter than the minimum length will be ignored.

See also

set_min_length()

Returns:

int Minimum String Length

OCRStringTemplate &set_min_length(int min_length)

Set Minimum String Length with int value.

String shorter than the minimum length will be ignored.

See also

get_min_length()

Parameters:

min_length – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRStringTemplate& the reference of this object.

double get_max_char_interval() const

Get Maximum Character Interval.

The maximum interval between characters in a string, Two characters with interval longer than this threshold will be matched into two different strings.

See also

set_max_char_interval()

Returns:

double Maximum Character Interval

OCRStringTemplate &set_max_char_interval(double max_char_interval)

Set Maximum Character Interval with double value.

The maximum interval between characters in a string, Two characters with interval longer than this threshold will be matched into two different strings.

See also

get_max_char_interval()

Parameters:

max_char_interval – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRStringTemplate& the reference of this object.

int get_main_direction() const

Get String Main Direction.

Direction of strings based on statistics. (It is specified that the positive x-axis direction is 0 degrees, positive values representing the direction towards the positive y-axis, negative values indicating the direction towards the negative y-axis.)

See also

set_main_direction()

Returns:

int String Main Direction

OCRStringTemplate &set_main_direction(int main_direction)

Set String Main Direction with int value.

Direction of strings based on statistics. (It is specified that the positive x-axis direction is 0 degrees, positive values representing the direction towards the positive y-axis, negative values indicating the direction towards the negative y-axis.)

See also

get_main_direction()

Parameters:

main_direction – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRStringTemplate& the reference of this object.

int get_angle_range() const

Get Rotate Angle Range.

The rotate angle range of the string, which means that the string can be rotated slightly within an angular range centered on the main direction.

See also

set_angle_range()

Returns:

int Rotate Angle Range

OCRStringTemplate &set_angle_range(int angle_range)

Set Rotate Angle Range with int value.

The rotate angle range of the string, which means that the string can be rotated slightly within an angular range centered on the main direction.

See also

get_angle_range()

Parameters:

angle_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRStringTemplate& the reference of this object.

const RegularExpression &get_regular_expression() const

Get Regular Expression.

Regular expression parameter.

See also

set_regular_expression()

Returns:

const RegularExpression & Regular Expression

OCRStringTemplate &set_regular_expression(const RegularExpression &regular_expression)

Set Regular Expression with RegularExpression value.

Regular expression parameter.

See also

get_regular_expression()

Parameters:

regular_expression – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRStringTemplate& the reference of this object.

RegularExpression &get_regular_expression()

Get mutable reference of Regular Expression.

Regular expression parameter.

Returns:

OCRStringTemplate& the mutable reference of the group.

class OCRFilterParameters : public visionflow::param::SchemableParameter

OCRFilterParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

double get_threshold() const

Get Confidence Threshold.

Greater than this threshold will be determined as a target.

See also

set_threshold()

Returns:

double Confidence Threshold

OCRFilterParameters &set_threshold(double threshold)

Set Confidence Threshold with double value.

Greater than this threshold will be determined as a target.

See also

get_threshold()

Parameters:

threshold – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRFilterParameters& the reference of this object.

double get_density() const

Get Search Density.

Filter out duplicate results using NMS. A higher value will retain fewer results.

See also

set_density()

Returns:

double Search Density

OCRFilterParameters &set_density(double density)

Set Search Density with double value.

Filter out duplicate results using NMS. A higher value will retain fewer results.

See also

get_density()

Parameters:

density – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRFilterParameters& the reference of this object.

class OCRInferParameters : public visionflow::param::SchemableParameter

OCRInferParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Customize Character Size.

Enable this option to customize target character size if the character in inference image is not match with the size of training data.

See also

set_enable()

Returns:

bool Customize Character Size

OCRInferParameters &set_enable(bool enable)

Set Customize Character Size with bool value.

Enable this option to customize target character size if the character in inference image is not match with the size of training data.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRInferParameters& the reference of this object.

double get_target_feat_width() const

Get Target Character Width.

The target character width in inference data.

See also

set_target_feat_width()

Returns:

double Target Character Width

OCRInferParameters &set_target_feat_width(double target_feat_width)

Set Target Character Width with double value.

The target character width in inference data.

See also

get_target_feat_width()

Parameters:

target_feat_width – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRInferParameters& the reference of this object.

double get_target_feat_height() const

Get Target Character Height.

The target character height in inference data.

See also

set_target_feat_height()

Returns:

double Target Character Height

OCRInferParameters &set_target_feat_height(double target_feat_height)

Set Target Character Height with double value.

The target character height in inference data.

See also

get_target_feat_height()

Parameters:

target_feat_height – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRInferParameters& the reference of this object.

class OCRUniversalModelParameters : public visionflow::param::SchemableParameter

OCRUniversalModelParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

double get_target_feat_width() const

Get Target Character Width.

The target character width in inference data.

See also

set_target_feat_width()

Returns:

double Target Character Width

OCRUniversalModelParameters &set_target_feat_width(double target_feat_width)

Set Target Character Width with double value.

The target character width in inference data.

See also

get_target_feat_width()

Parameters:

target_feat_width – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRUniversalModelParameters& the reference of this object.

double get_target_feat_height() const

Get Target Character Height.

The target character height in inference data.

See also

set_target_feat_height()

Returns:

double Target Character Height

OCRUniversalModelParameters &set_target_feat_height(double target_feat_height)

Set Target Character Height with double value.

The target character height in inference data.

See also

get_target_feat_height()

Parameters:

target_feat_height – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRUniversalModelParameters& the reference of this object.

const OCRAnnulusParameters &get_annulus_recognition() const

Get Annulus Character Recognition.

Annulus Character Recognition Parameters

See also

set_annulus_recognition()

Returns:

const OCRAnnulusParameters & Annulus Character Recognition

OCRUniversalModelParameters &set_annulus_recognition(const OCRAnnulusParameters &annulus_recognition)

Set Annulus Character Recognition with OCRAnnulusParameters value.

Annulus Character Recognition Parameters

See also

get_annulus_recognition()

Parameters:

annulus_recognition – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRUniversalModelParameters& the reference of this object.

OCRAnnulusParameters &get_annulus_recognition()

Get mutable reference of Annulus Character Recognition.

Annulus Character Recognition Parameters

Returns:

OCRUniversalModelParameters& the mutable reference of the group.

class OCRTemplates : public visionflow::param::SchemableParameter

OCRTemplates Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const std::map<std::string, OCRStringTemplate> &get_string_templates() const

Get String Templates.

String templates.

See also

set_string_templates()

Returns:

const std::map<std::string, OCRStringTemplate> & String Templates

OCRTemplates &set_string_templates(const std::map<std::string, OCRStringTemplate> &string_templates)

Set String Templates with std::map<std::string, OCRStringTemplate> value.

String templates.

See also

get_string_templates()

Parameters:

string_templates – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTemplates& the reference of this object.

const OCRStringTemplate &get_string_templates(const std::string &key) const

Get value in String Templates with key.

Warning

The key must be exist in String Templates. otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const std::map<std::string, OCRStringTemplate> & value in String Templates at key.

OCRTemplates &set_string_templates(const std::string &key, const OCRStringTemplate &value)

Set value in String Templates with key.

Parameters:

• key – the key to set.

• value – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTemplates& the reference of this object.

bool string_templates_contains(const std::string &key) const

Check if the key is exist in String Templates.

Returns:

bool true if the key is exist in String Templates, otherwise false.

size_t get_string_templates_size() const

Get the size of String Templates.

Returns:

size_t the size of String Templates

OCRStringTemplate &get_string_templates(const std::string &key)

Get mutable reference of value in String Templates with key. A new key and default value will be created if the key does not exist.

Returns:

OCRStringTemplate& the mutable reference of value in String Templates at key.

const std::map<std::string, OCRNodeTemplate> &get_node_templates() const

Get Node Templates.

Node templates.

See also

set_node_templates()

Returns:

const std::map<std::string, OCRNodeTemplate> & Node Templates

OCRTemplates &set_node_templates(const std::map<std::string, OCRNodeTemplate> &node_templates)

Set Node Templates with std::map<std::string, OCRNodeTemplate> value.

Node templates.

See also

get_node_templates()

Parameters:

node_templates – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTemplates& the reference of this object.

const OCRNodeTemplate &get_node_templates(const std::string &key) const

Get value in Node Templates with key.

Warning

The key must be exist in Node Templates. otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const std::map<std::string, OCRNodeTemplate> & value in Node Templates at key.

OCRTemplates &set_node_templates(const std::string &key, const OCRNodeTemplate &value)

Set value in Node Templates with key.

Parameters:

• key – the key to set.

• value – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTemplates& the reference of this object.

bool node_templates_contains(const std::string &key) const

Check if the key is exist in Node Templates.

Returns:

bool true if the key is exist in Node Templates, otherwise false.

size_t get_node_templates_size() const

Get the size of Node Templates.

Returns:

size_t the size of Node Templates

OCRNodeTemplate &get_node_templates(const std::string &key)

Get mutable reference of value in Node Templates with key. A new key and default value will be created if the key does not exist.

Returns:

OCRNodeTemplate& the mutable reference of value in Node Templates at key.

class OCRAnnulusParameters : public visionflow::param::ISchemable

OCRAnnulusParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Annulus Recognition.

Enable this option to recognize characters arranged in a circular pattern.

See also

set_enable()

Returns:

bool Enable Annulus Recognition

OCRAnnulusParameters &set_enable(bool enable)

Set Enable Annulus Recognition with bool value.

Enable this option to recognize characters arranged in a circular pattern.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRAnnulusParameters& the reference of this object.

double get_x() const

Get Center X.

The x-coordinate of the center of the annulus, set according to the percentage of the view height.

See also

set_x()

Returns:

double Center X

OCRAnnulusParameters &set_x(double x)

Set Center X with double value.

The x-coordinate of the center of the annulus, set according to the percentage of the view height.

See also

get_x()

Parameters:

x – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRAnnulusParameters& the reference of this object.

double get_y() const

Get Center Y.

The y-coordinate of the center of the annulus, set according to the percentage of the view height.

See also

set_y()

Returns:

double Center Y

OCRAnnulusParameters &set_y(double y)

Set Center Y with double value.

The y-coordinate of the center of the annulus, set according to the percentage of the view height.

See also

get_y()

Parameters:

y – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRAnnulusParameters& the reference of this object.

double get_inner_radius() const

Get Inner Radius.

The inner radius of the annulus, set according to the percentage of the view height.

See also

set_inner_radius()

Returns:

double Inner Radius

OCRAnnulusParameters &set_inner_radius(double inner_radius)

Set Inner Radius with double value.

The inner radius of the annulus, set according to the percentage of the view height.

See also

get_inner_radius()

Parameters:

inner_radius – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRAnnulusParameters& the reference of this object.

double get_outer_radius() const

Get Outer Radius.

The outer radius of the annulus, set according to the percentage of the view height.

See also

set_outer_radius()

Returns:

double Outer Radius

OCRAnnulusParameters &set_outer_radius(double outer_radius)

Set Outer Radius with double value.

The outer radius of the annulus, set according to the percentage of the view height.

See also

get_outer_radius()

Parameters:

outer_radius – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRAnnulusParameters& the reference of this object.

double get_start_angle() const

Get Start Angle.

It is specified that the positive x-axis direction is 0 degrees, positive values representing the direction towards the negative y-axis, negative values indicating the direction towards the positive y-axis.

See also

set_start_angle()

Returns:

double Start Angle

OCRAnnulusParameters &set_start_angle(double start_angle)

Set Start Angle with double value.

It is specified that the positive x-axis direction is 0 degrees, positive values representing the direction towards the negative y-axis, negative values indicating the direction towards the positive y-axis.

See also

get_start_angle()

Parameters:

start_angle – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRAnnulusParameters& the reference of this object.

double get_angle_range() const

Get Angle Range.

The range of angle in which the annulus is unwrapped, if the angle is greater than 360 degrees or less than -360 degrees, the excess part will be unwrapped repeatedly. Positive values representing the direction towards the negative y-axis, negative values indicating the direction towards the positive y-axis.

See also

set_angle_range()

Returns:

double Angle Range

OCRAnnulusParameters &set_angle_range(double angle_range)

Set Angle Range with double value.

The range of angle in which the annulus is unwrapped, if the angle is greater than 360 degrees or less than -360 degrees, the excess part will be unwrapped repeatedly. Positive values representing the direction towards the negative y-axis, negative values indicating the direction towards the positive y-axis.

See also

get_angle_range()

Parameters:

angle_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRAnnulusParameters& the reference of this object.

class OCRTrainingParameters : public visionflow::param::SchemableParameter

OCRTrainingParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int get_epoch() const

Get Epochs.

Each iteration of the training images participates in a training session.

See also

set_epoch()

Returns:

int Epochs

OCRTrainingParameters &set_epoch(int epoch)

Set Epochs with int value.

Each iteration of the training images participates in a training session.

See also

get_epoch()

Parameters:

epoch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTrainingParameters& the reference of this object.

int get_batch_size() const

Get Training Batch Size.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16.

See also

set_batch_size()

Returns:

int Training Batch Size

OCRTrainingParameters &set_batch_size(int batch_size)

Set Training Batch Size with int value.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16.

See also

get_batch_size()

Parameters:

batch_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTrainingParameters& the reference of this object.

const TrainingMode &get_training_mode() const

Get Training Mode.

kNormalTrain stands for the normal training mode, in which the model is trained from scratch, ignoring any pre-existing training results. kIncrementalTrain stands for the incremental training mode, where the model continues training on top of the pre-existing training results, saving time and making full use of the existing data. By default, the incremental training mode is used.

See also

set_training_mode()

Returns:

const TrainingMode & Training Mode

OCRTrainingParameters &set_training_mode(const TrainingMode &training_mode)

Set Training Mode with TrainingMode value.

kNormalTrain stands for the normal training mode, in which the model is trained from scratch, ignoring any pre-existing training results. kIncrementalTrain stands for the incremental training mode, where the model continues training on top of the pre-existing training results, saving time and making full use of the existing data. By default, the incremental training mode is used.

See also

get_training_mode()

Parameters:

training_mode – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTrainingParameters& the reference of this object.

double get_sampling_ratio() const

Get Sampling Ratio.

The proportion of new data in incremental training. This parameter is only applicable when the training mode is set to ‘kIncrementalTrain’.

See also

set_sampling_ratio()

Returns:

double Sampling Ratio

OCRTrainingParameters &set_sampling_ratio(double sampling_ratio)

Set Sampling Ratio with double value.

The proportion of new data in incremental training. This parameter is only applicable when the training mode is set to ‘kIncrementalTrain’.

See also

get_sampling_ratio()

Parameters:

sampling_ratio – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTrainingParameters& the reference of this object.

bool get_only_check() const

Get Only Check.

Default is False. If set to True, the trainer will only check if it is possible to train in the given training mode without actually performing the training.

See also

set_only_check()

Returns:

bool Only Check

OCRTrainingParameters &set_only_check(bool only_check)

Set Only Check with bool value.

Default is False. If set to True, the trainer will only check if it is possible to train in the given training mode without actually performing the training.

See also

get_only_check()

Parameters:

only_check – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTrainingParameters& the reference of this object.

double get_step_ratio() const

Get Learning Rate Update Step Ratio.

The ratio of iterations after which the learning rate is decreased.

See also

set_step_ratio()

Returns:

double Learning Rate Update Step Ratio

OCRTrainingParameters &set_step_ratio(double step_ratio)

Set Learning Rate Update Step Ratio with double value.

The ratio of iterations after which the learning rate is decreased.

See also

get_step_ratio()

Parameters:

step_ratio – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTrainingParameters& the reference of this object.

int get_sigma() const

Get Gaussian Kernel Size.

Determines the spread of Gaussian distribution used to generate heatmap. A larger value will result in a larger area of distribution, vice versa.

See also

set_sigma()

Returns:

int Gaussian Kernel Size

OCRTrainingParameters &set_sigma(int sigma)

Set Gaussian Kernel Size with int value.

Determines the spread of Gaussian distribution used to generate heatmap. A larger value will result in a larger area of distribution, vice versa.

See also

get_sigma()

Parameters:

sigma – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTrainingParameters& the reference of this object.

int get_training_feature_size() const

Get Training Feature Size.

The desired size to which the input features are scaled

See also

set_training_feature_size()

Returns:

int Training Feature Size

OCRTrainingParameters &set_training_feature_size(int training_feature_size)

Set Training Feature Size with int value.

The desired size to which the input features are scaled

See also

get_training_feature_size()

Parameters:

training_feature_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTrainingParameters& the reference of this object.

const OCRAnnulusParameters &get_annulus_recognition() const

Get Annulus Character Recognition.

Annulus Character Recognition Parameters

See also

set_annulus_recognition()

Returns:

const OCRAnnulusParameters & Annulus Character Recognition

OCRTrainingParameters &set_annulus_recognition(const OCRAnnulusParameters &annulus_recognition)

Set Annulus Character Recognition with OCRAnnulusParameters value.

Annulus Character Recognition Parameters

See also

get_annulus_recognition()

Parameters:

annulus_recognition – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTrainingParameters& the reference of this object.

OCRAnnulusParameters &get_annulus_recognition()

Get mutable reference of Annulus Character Recognition.

Annulus Character Recognition Parameters

Returns:

OCRTrainingParameters& the mutable reference of the group.

const DataAugmentation &get_augmentations() const

Get Data Augmentation.

Data Augmentation Parameters

See also

set_augmentations()

Returns:

const DataAugmentation & Data Augmentation

OCRTrainingParameters &set_augmentations(const DataAugmentation &augmentations)

Set Data Augmentation with DataAugmentation value.

Data Augmentation Parameters

See also

get_augmentations()

Parameters:

augmentations – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

OCRTrainingParameters& the reference of this object.

DataAugmentation &get_augmentations()

Get mutable reference of Data Augmentation.

Data Augmentation Parameters

Returns:

OCRTrainingParameters& the mutable reference of the group.

Location Tool Parameters

class LocationMaxInputSize : public visionflow::param::ISchemable

LocationMaxInputSize Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Set Max Size.

Enable this option to manually set the maximum side length of the input image.

See also

set_enable()

Returns:

bool Enable Set Max Size

LocationMaxInputSize &set_enable(bool enable)

Set Enable Set Max Size with bool value.

Enable this option to manually set the maximum side length of the input image.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationMaxInputSize& the reference of this object.

int get_max_side_len() const

Get Max Input Size.

Scale the long edge of the input image to the maximum side length, and scale the short edge equally, if the algorithm calculates that the required maximum edge length is smaller than this parameter, the maximum edge length will be set automatically

See also

set_max_side_len()

Returns:

int Max Input Size

LocationMaxInputSize &set_max_side_len(int max_side_len)

Set Max Input Size with int value.

Scale the long edge of the input image to the maximum side length, and scale the short edge equally, if the algorithm calculates that the required maximum edge length is smaller than this parameter, the maximum edge length will be set automatically

See also

get_max_side_len()

Parameters:

max_side_len – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationMaxInputSize& the reference of this object.

class LocationModelParameters : public visionflow::param::ISchemable

LocationModelParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const std::string &get_model_arch() const

Get Model Architecture.

There are two different model architectures: the High-Precision-Location model and Fast-Location model, The high-precision version is suitable for scenarios with positioning accuracy up to 1 pixel, and the fast version is suitable for most scenarios with shorter training, inference time and lower GPU memory requirements.

See also

set_model_arch()

Returns:

const std::string & Model Architecture

LocationModelParameters &set_model_arch(const std::string &model_arch)

Set Model Architecture with std::string value.

There are two different model architectures: the High-Precision-Location model and Fast-Location model, The high-precision version is suitable for scenarios with positioning accuracy up to 1 pixel, and the fast version is suitable for most scenarios with shorter training, inference time and lower GPU memory requirements.

See also

get_model_arch()

Parameters:

model_arch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationModelParameters& the reference of this object.

bool get_train_angle() const

Get Training With Angle.

Enable this option to training the model with the angle information in the label.

See also

set_train_angle()

Returns:

bool Training With Angle

LocationModelParameters &set_train_angle(bool train_angle)

Set Training With Angle with bool value.

Enable this option to training the model with the angle information in the label.

See also

get_train_angle()

Parameters:

train_angle – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationModelParameters& the reference of this object.

bool get_train_rect() const

Get Training With Rect Size.

Enable this option to training the model with the the rectangle size information of the target box

See also

set_train_rect()

Returns:

bool Training With Rect Size

LocationModelParameters &set_train_rect(bool train_rect)

Set Training With Rect Size with bool value.

Enable this option to training the model with the the rectangle size information of the target box

See also

get_train_rect()

Parameters:

train_rect – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationModelParameters& the reference of this object.

class LocationTargetParameters : public visionflow::param::ISchemable

LocationTargetParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

double get_min_target_width() const

Get Minimum Target Width.

The Minimum target width (Pixel value in X-axis direction)

See also

set_min_target_width()

Returns:

double Minimum Target Width

LocationTargetParameters &set_min_target_width(double min_target_width)

Set Minimum Target Width with double value.

The Minimum target width (Pixel value in X-axis direction)

See also

get_min_target_width()

Parameters:

min_target_width – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationTargetParameters& the reference of this object.

double get_min_target_height() const

Get Minimum Target Height.

The Minimum target height (Pixel value in Y-axis direction)

See also

set_min_target_height()

Returns:

double Minimum Target Height

LocationTargetParameters &set_min_target_height(double min_target_height)

Set Minimum Target Height with double value.

The Minimum target height (Pixel value in Y-axis direction)

See also

get_min_target_height()

Parameters:

min_target_height – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationTargetParameters& the reference of this object.

class NodeMatchTemplate : public visionflow::param::ISchemable

NodeMatchTemplate Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Note Match Template.

Enable this option to use this template

See also

set_enable()

Returns:

bool Enable Note Match Template

NodeMatchTemplate &set_enable(bool enable)

Set Enable Note Match Template with bool value.

Enable this option to use this template

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

const TemplateMode &get_template_mode() const

Get Template Mode.

Set the value type of points and offset. When the template mode is set to kValueByPixel, the Coordinates of points and the offset should be set using pixel value. When the template mode is set to kValueByRatio, the Coordinates of points and the offset should be set using ratio value. When the Area Value Type is modified, the Coordinates of points and the offset should be modified.

See also

set_template_mode()

Returns:

const TemplateMode & Template Mode

NodeMatchTemplate &set_template_mode(const TemplateMode &template_mode)

Set Template Mode with TemplateMode value.

Set the value type of points and offset. When the template mode is set to kValueByPixel, the Coordinates of points and the offset should be set using pixel value. When the template mode is set to kValueByRatio, the Coordinates of points and the offset should be set using ratio value. When the Area Value Type is modified, the Coordinates of points and the offset should be modified.

See also

get_template_mode()

Parameters:

template_mode – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

int get_max_match_results_num() const

Get Max Match Results Num.

the maximum number of template matching results. When the value is - 1, the maximum number of matching results is not limited.

See also

set_max_match_results_num()

Returns:

int Max Match Results Num

NodeMatchTemplate &set_max_match_results_num(int max_match_results_num)

Set Max Match Results Num with int value.

the maximum number of template matching results. When the value is - 1, the maximum number of matching results is not limited.

See also

get_max_match_results_num()

Parameters:

max_match_results_num – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

int get_max_miss_points_num() const

Get Max Miss Points Num.

The number of points allowed to be missed during matching

See also

set_max_miss_points_num()

Returns:

int Max Miss Points Num

NodeMatchTemplate &set_max_miss_points_num(int max_miss_points_num)

Set Max Miss Points Num with int value.

The number of points allowed to be missed during matching

See also

get_max_miss_points_num()

Parameters:

max_miss_points_num – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

const std::vector<KeyPointNode> &get_points() const

Get Keypoints.

The key points of the template

See also

set_points()

Returns:

const std::vector<KeyPointNode> & Keypoints

NodeMatchTemplate &set_points(const std::vector<KeyPointNode> &points)

Set Keypoints with std::vector<KeyPointNode> value.

The key points of the template

See also

get_points()

Parameters:

points – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

const KeyPointNode &get_points(size_t index) const

Get value in Keypoints with index.

Warning

The index must be less than get_points_size(). otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const KeyPointNode & value in Keypoints at index.

size_t get_points_size() const

Get the size of Keypoints.

Returns:

size_t the size of Keypoints

KeyPointNode &get_points(size_t index)

Get mutable reference of value in Keypoints with index.

Warning

The index must be less than get_points_size(). otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

KeyPointNode& the mutable reference of value in Keypoints at index.

double get_max_distance() const

Get Max Distance.

The maximum allowed distance between the template keypoints and the keypoints to be matched. Results with a distance greater than this value will be filtered out. The value can be either in absolute pixel or relative to the view height, depending on the template_mode.

See also

set_max_distance()

Returns:

double Max Distance

NodeMatchTemplate &set_max_distance(double max_distance)

Set Max Distance with double value.

The maximum allowed distance between the template keypoints and the keypoints to be matched. Results with a distance greater than this value will be filtered out. The value can be either in absolute pixel or relative to the view height, depending on the template_mode.

See also

get_max_distance()

Parameters:

max_distance – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

bool get_is_scalable() const

Get Target Scalable.

Enable this option if the scale of target to be matched is not fixed, And you should set the target scale range

See also

set_is_scalable()

Returns:

bool Target Scalable

NodeMatchTemplate &set_is_scalable(bool is_scalable)

Set Target Scalable with bool value.

Enable this option if the scale of target to be matched is not fixed, And you should set the target scale range

See also

get_is_scalable()

Parameters:

is_scalable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

const std::vector<double> &get_scale_range() const

Get Scale Range.

The target scale range

See also

set_scale_range()

Returns:

const std::vector<double> & Scale Range

NodeMatchTemplate &set_scale_range(const std::vector<double> &scale_range)

Set Scale Range with std::vector<double> value.

The target scale range

See also

get_scale_range()

Parameters:

scale_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

bool scale_range_contains(double value) const

Check if the Scale Range contains the value.

Returns:

bool true if the Scale Range contains the value, otherwise false.

double get_scale_range_left() const

Get left point value of Scale Range.

See also

set_scale_range_right()

Returns:

const std::vector<double> & left point value of Scale Range

double get_scale_range_right() const

Get the right point value of Scale Range.

See also

set_scale_range_left()

Returns:

const std::vector<double> & the right point value of Scale Range

NodeMatchTemplate &set_scale_range_left(double scale_range_left)

Set left point value of Scale Range with std::vector<double> value.

The target scale range

See also

get_scale_range_left()

Parameters:

scale_range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

NodeMatchTemplate &set_scale_range_right(double scale_range_right)

Set the right point value of Scale Range with std::vector<double> value.

The target scale range

See also

get_scale_range_right()

Parameters:

scale_range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

bool get_is_rotatable() const

Get Target Rotatable.

Enable this option if the angle of target to be matched is not fixed, And you should set the target rotate angle range

See also

set_is_rotatable()

Returns:

bool Target Rotatable

NodeMatchTemplate &set_is_rotatable(bool is_rotatable)

Set Target Rotatable with bool value.

Enable this option if the angle of target to be matched is not fixed, And you should set the target rotate angle range

See also

get_is_rotatable()

Parameters:

is_rotatable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

const std::vector<double> &get_rotate_range() const

Get Rotate Range (Degree).

The target rotate range

See also

set_rotate_range()

Returns:

const std::vector<double> & Rotate Range (Degree)

NodeMatchTemplate &set_rotate_range(const std::vector<double> &rotate_range)

Set Rotate Range (Degree) with std::vector<double> value.

The target rotate range

See also

get_rotate_range()

Parameters:

rotate_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

bool rotate_range_contains(double value) const

Check if the Rotate Range (Degree) contains the value.

Returns:

bool true if the Rotate Range (Degree) contains the value, otherwise false.

double get_rotate_range_left() const

Get left point value of Rotate Range (Degree).

See also

set_rotate_range_right()

Returns:

const std::vector<double> & left point value of Rotate Range (Degree)

double get_rotate_range_right() const

Get the right point value of Rotate Range (Degree).

See also

set_rotate_range_left()

Returns:

const std::vector<double> & the right point value of Rotate Range (Degree)

NodeMatchTemplate &set_rotate_range_left(double rotate_range_left)

Set left point value of Rotate Range (Degree) with std::vector<double> value.

The target rotate range

See also

get_rotate_range_left()

Parameters:

rotate_range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

NodeMatchTemplate &set_rotate_range_right(double rotate_range_right)

Set the right point value of Rotate Range (Degree) with std::vector<double> value.

The target rotate range

See also

get_rotate_range_right()

Parameters:

rotate_range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

double get_template_angle() const

Get Template angle (Degree).

Template angle is calculated and displayed the software in real time

See also

set_template_angle()

Returns:

double Template angle (Degree)

NodeMatchTemplate &set_template_angle(double template_angle)

Set Template angle (Degree) with double value.

Template angle is calculated and displayed the software in real time

See also

get_template_angle()

Parameters:

template_angle – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

NodeMatchTemplate& the reference of this object.

class LocationTemplates : public visionflow::param::SchemableParameter

LocationTemplates Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const std::map<std::string, NodeMatchTemplate> &get_node_templates() const

Get Node Templates.

The Location node match templates

See also

set_node_templates()

Returns:

const std::map<std::string, NodeMatchTemplate> & Node Templates

LocationTemplates &set_node_templates(const std::map<std::string, NodeMatchTemplate> &node_templates)

Set Node Templates with std::map<std::string, NodeMatchTemplate> value.

The Location node match templates

See also

get_node_templates()

Parameters:

node_templates – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationTemplates& the reference of this object.

const NodeMatchTemplate &get_node_templates(const std::string &key) const

Get value in Node Templates with key.

Warning

The key must be exist in Node Templates. otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const std::map<std::string, NodeMatchTemplate> & value in Node Templates at key.

LocationTemplates &set_node_templates(const std::string &key, const NodeMatchTemplate &value)

Set value in Node Templates with key.

Parameters:

• key – the key to set.

• value – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationTemplates& the reference of this object.

bool node_templates_contains(const std::string &key) const

Check if the key is exist in Node Templates.

Returns:

bool true if the key is exist in Node Templates, otherwise false.

size_t get_node_templates_size() const

Get the size of Node Templates.

Returns:

size_t the size of Node Templates

NodeMatchTemplate &get_node_templates(const std::string &key)

Get mutable reference of value in Node Templates with key. A new key and default value will be created if the key does not exist.

Returns:

NodeMatchTemplate& the mutable reference of value in Node Templates at key.

class LocationTrainingParameters : public visionflow::param::SchemableParameter

LocationTrainingParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int get_epoch() const

Get Epoch.

Number of epochs to train

See also

set_epoch()

Returns:

int Epoch

LocationTrainingParameters &set_epoch(int epoch)

Set Epoch with int value.

Number of epochs to train

See also

get_epoch()

Parameters:

epoch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationTrainingParameters& the reference of this object.

int get_batch_size() const

Get Training Batch Size.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16. The classification module needs to be set larger, generally 32, 64

See also

set_batch_size()

Returns:

int Training Batch Size

LocationTrainingParameters &set_batch_size(int batch_size)

Set Training Batch Size with int value.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16. The classification module needs to be set larger, generally 32, 64

See also

get_batch_size()

Parameters:

batch_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationTrainingParameters& the reference of this object.

const DataAugmentation &get_augmentations() const

Get Data Augmentation.

Data Augmentation Parameters

See also

set_augmentations()

Returns:

const DataAugmentation & Data Augmentation

LocationTrainingParameters &set_augmentations(const DataAugmentation &augmentations)

Set Data Augmentation with DataAugmentation value.

Data Augmentation Parameters

See also

get_augmentations()

Parameters:

augmentations – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationTrainingParameters& the reference of this object.

DataAugmentation &get_augmentations()

Get mutable reference of Data Augmentation.

Data Augmentation Parameters

Returns:

LocationTrainingParameters& the mutable reference of the group.

const LocationMaxInputSize &get_input_shape() const

Get Input Shape.

Input Shape Parameters

See also

set_input_shape()

Returns:

const LocationMaxInputSize & Input Shape

LocationTrainingParameters &set_input_shape(const LocationMaxInputSize &input_shape)

Set Input Shape with LocationMaxInputSize value.

Input Shape Parameters

See also

get_input_shape()

Parameters:

input_shape – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationTrainingParameters& the reference of this object.

LocationMaxInputSize &get_input_shape()

Get mutable reference of Input Shape.

Input Shape Parameters

Returns:

LocationTrainingParameters& the mutable reference of the group.

const LocationModelParameters &get_model_param() const

Get Model param.

Location model parameters

See also

set_model_param()

Returns:

const LocationModelParameters & Model param

LocationTrainingParameters &set_model_param(const LocationModelParameters &model_param)

Set Model param with LocationModelParameters value.

Location model parameters

See also

get_model_param()

Parameters:

model_param – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationTrainingParameters& the reference of this object.

LocationModelParameters &get_model_param()

Get mutable reference of Model param.

Location model parameters

Returns:

LocationTrainingParameters& the mutable reference of the group.

const LocationTargetParameters &get_target_feature() const

Get Target Feature.

Location Target Feature Parameters

See also

set_target_feature()

Returns:

const LocationTargetParameters & Target Feature

LocationTrainingParameters &set_target_feature(const LocationTargetParameters &target_feature)

Set Target Feature with LocationTargetParameters value.

Location Target Feature Parameters

See also

get_target_feature()

Parameters:

target_feature – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationTrainingParameters& the reference of this object.

LocationTargetParameters &get_target_feature()

Get mutable reference of Target Feature.

Location Target Feature Parameters

Returns:

LocationTrainingParameters& the mutable reference of the group.

class LocationFilterParameters : public visionflow::param::SchemableParameter

LocationFilterParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

double get_threshold() const

Get Confidence Threshold.

Keypoints with confidence greater than this threshold will be determined as a character target.

See also

set_threshold()

Returns:

double Confidence Threshold

LocationFilterParameters &set_threshold(double threshold)

Set Confidence Threshold with double value.

Keypoints with confidence greater than this threshold will be determined as a character target.

See also

get_threshold()

Parameters:

threshold – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationFilterParameters& the reference of this object.

double get_target_overlap_ratio() const

Get Target Overlap Ratio.

Some redundant points will be removed to ensure that the intersection ratio between points within each category does not exceed this value.target_overlap_ratio = intersection_area/Union_area.

See also

set_target_overlap_ratio()

Returns:

double Target Overlap Ratio

LocationFilterParameters &set_target_overlap_ratio(double target_overlap_ratio)

Set Target Overlap Ratio with double value.

Some redundant points will be removed to ensure that the intersection ratio between points within each category does not exceed this value.target_overlap_ratio = intersection_area/Union_area.

See also

get_target_overlap_ratio()

Parameters:

target_overlap_ratio – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

LocationFilterParameters& the reference of this object.

AssemblyVerification Tool Parameters

class AssemblyVerificationMaxInputSize : public visionflow::param::ISchemable

AssemblyVerificationMaxInputSize Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Set Max Size.

Enable this option to manually set the maximum side length of the input image.

See also

set_enable()

Returns:

bool Enable Set Max Size

AssemblyVerificationMaxInputSize &set_enable(bool enable)

Set Enable Set Max Size with bool value.

Enable this option to manually set the maximum side length of the input image.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationMaxInputSize& the reference of this object.

int get_max_side_len() const

Get Max Input Size.

Scale the long edge of the input image to the maximum side length, and scale the short edge equally, if the algorithm calculates that the required maximum edge length is smaller than this parameter, the maximum edge length will be set automatically

See also

set_max_side_len()

Returns:

int Max Input Size

AssemblyVerificationMaxInputSize &set_max_side_len(int max_side_len)

Set Max Input Size with int value.

Scale the long edge of the input image to the maximum side length, and scale the short edge equally, if the algorithm calculates that the required maximum edge length is smaller than this parameter, the maximum edge length will be set automatically

See also

get_max_side_len()

Parameters:

max_side_len – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationMaxInputSize& the reference of this object.

class AssemblyVerificationModelParameters : public visionflow::param::ISchemable

AssemblyVerificationModelParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_train_angle() const

Get Training With Angle.

Enable this option to training the model with the angle infomations in the label.

See also

set_train_angle()

Returns:

bool Training With Angle

AssemblyVerificationModelParameters &set_train_angle(bool train_angle)

Set Training With Angle with bool value.

Enable this option to training the model with the angle infomations in the label.

See also

get_train_angle()

Parameters:

train_angle – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationModelParameters& the reference of this object.

bool get_train_rect() const

Get Training With Rect Size.

Enable this option to training the model with the the rectangle size infomation of the target box

See also

set_train_rect()

Returns:

bool Training With Rect Size

AssemblyVerificationModelParameters &set_train_rect(bool train_rect)

Set Training With Rect Size with bool value.

Enable this option to training the model with the the rectangle size infomation of the target box

See also

get_train_rect()

Parameters:

train_rect – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationModelParameters& the reference of this object.

class AssemblyVerificationTargetParameters : public visionflow::param::ISchemable

AssemblyVerificationTargetParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

double get_min_target_width() const

Get Minimum Target Width.

The Minimum target width (Pixel value in X-axis direction)

See also

set_min_target_width()

Returns:

double Minimum Target Width

AssemblyVerificationTargetParameters &set_min_target_width(double min_target_width)

Set Minimum Target Width with double value.

The Minimum target width (Pixel value in X-axis direction)

See also

get_min_target_width()

Parameters:

min_target_width – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationTargetParameters& the reference of this object.

double get_min_target_height() const

Get Minimum Target Height.

The Minimum target height (Pixel value in Y-axis direction)

See also

set_min_target_height()

Returns:

double Minimum Target Height

AssemblyVerificationTargetParameters &set_min_target_height(double min_target_height)

Set Minimum Target Height with double value.

The Minimum target height (Pixel value in Y-axis direction)

See also

get_min_target_height()

Parameters:

min_target_height – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationTargetParameters& the reference of this object.

class AssemblyVerificationLayOutArea : public visionflow::param::ISchemable

AssemblyVerificationLayOutArea Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

double get_x() const

Get X-Value.

X-direction coordinate value of the center of the area. When the template mode is set to kValueByPixel, X-Value should be set using pixel value. When the template mode is set to kValueByRatio, X-Value should be set using ratio value. When the template mode is modified, X-Value needs to be modified.

See also

set_x()

Returns:

double X-Value

AssemblyVerificationLayOutArea &set_x(double x)

Set X-Value with double value.

X-direction coordinate value of the center of the area. When the template mode is set to kValueByPixel, X-Value should be set using pixel value. When the template mode is set to kValueByRatio, X-Value should be set using ratio value. When the template mode is modified, X-Value needs to be modified.

See also

get_x()

Parameters:

x – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationLayOutArea& the reference of this object.

double get_y() const

Get Y-Value.

Y-direction coordinate value of the center of the area. When the template mode is set to kValueByPixel, Y-Value should be set using pixel value. When the template mode is set to kValueByRatio, Y-Value should be set using ratio value. When the template mode is modified, Y-Value needs to be modified.

See also

set_y()

Returns:

double Y-Value

AssemblyVerificationLayOutArea &set_y(double y)

Set Y-Value with double value.

Y-direction coordinate value of the center of the area. When the template mode is set to kValueByPixel, Y-Value should be set using pixel value. When the template mode is set to kValueByRatio, Y-Value should be set using ratio value. When the template mode is modified, Y-Value needs to be modified.

See also

get_y()

Parameters:

y – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationLayOutArea& the reference of this object.

double get_w() const

Get W-Value.

width of the area. When the template mode is set to kValueByPixel, W-Value should be set using pixel value. When the template mode is set to kValueByRatio, W-Value should be set using ratio value. When the template mode is modified, W-Value needs to be modified.

See also

set_w()

Returns:

double W-Value

AssemblyVerificationLayOutArea &set_w(double w)

Set W-Value with double value.

width of the area. When the template mode is set to kValueByPixel, W-Value should be set using pixel value. When the template mode is set to kValueByRatio, W-Value should be set using ratio value. When the template mode is modified, W-Value needs to be modified.

See also

get_w()

Parameters:

w – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationLayOutArea& the reference of this object.

double get_h() const

Get H-Value.

height of the area. When the template mode is set to kValueByPixel, H-Value should be set using pixel value. When the template mode is set to kValueRatio, H-Value should be set using ratio value. When the template mode is modified, H-Value needs to be modified.

See also

set_h()

Returns:

double H-Value

AssemblyVerificationLayOutArea &set_h(double h)

Set H-Value with double value.

height of the area. When the template mode is set to kValueByPixel, H-Value should be set using pixel value. When the template mode is set to kValueRatio, H-Value should be set using ratio value. When the template mode is modified, H-Value needs to be modified.

See also

get_h()

Parameters:

h – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationLayOutArea& the reference of this object.

const std::vector<std::string> &get_target_category() const

Get Target Category.

Categories of targets in the area

See also

set_target_category()

Returns:

const std::vector<std::string> & Target Category

AssemblyVerificationLayOutArea &set_target_category(const std::vector<std::string> &target_category)

Set Target Category with std::vector<std::string> value.

Categories of targets in the area

See also

get_target_category()

Parameters:

target_category – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationLayOutArea& the reference of this object.

const std::string &get_target_category(size_t index) const

Get value in Target Category with index.

Warning

The index must be less than get_target_category_size(). otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const std::string & value in Target Category at index.

size_t get_target_category_size() const

Get the size of Target Category.

Returns:

size_t the size of Target Category

int get_target_number() const

Get Target Number.

Number of targets in the region

See also

set_target_number()

Returns:

int Target Number

AssemblyVerificationLayOutArea &set_target_number(int target_number)

Set Target Number with int value.

Number of targets in the region

See also

get_target_number()

Parameters:

target_number – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationLayOutArea& the reference of this object.

bool get_use_target_angle_range() const

Get Use Target Angle Range.

Enable this option if the angle of target to be matched needs to belimited, And you should set the target angle range.

See also

set_use_target_angle_range()

Returns:

bool Use Target Angle Range

AssemblyVerificationLayOutArea &set_use_target_angle_range(bool use_target_angle_range)

Set Use Target Angle Range with bool value.

Enable this option if the angle of target to be matched needs to belimited, And you should set the target angle range.

See also

get_use_target_angle_range()

Parameters:

use_target_angle_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationLayOutArea& the reference of this object.

const std::vector<double> &get_target_angle_range() const

Get Target Angle Range (Degree).

The angle range target limited to

See also

set_target_angle_range()

Returns:

const std::vector<double> & Target Angle Range (Degree)

AssemblyVerificationLayOutArea &set_target_angle_range(const std::vector<double> &target_angle_range)

Set Target Angle Range (Degree) with std::vector<double> value.

The angle range target limited to

See also

get_target_angle_range()

Parameters:

target_angle_range – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationLayOutArea& the reference of this object.

bool target_angle_range_contains(double value) const

Check if the Target Angle Range (Degree) contains the value.

Returns:

bool true if the Target Angle Range (Degree) contains the value, otherwise false.

double get_target_angle_range_left() const

Get left point value of Target Angle Range (Degree).

See also

set_target_angle_range_right()

Returns:

const std::vector<double> & left point value of Target Angle Range (Degree)

double get_target_angle_range_right() const

Get the right point value of Target Angle Range (Degree).

See also

set_target_angle_range_left()

Returns:

const std::vector<double> & the right point value of Target Angle Range (Degree)

AssemblyVerificationLayOutArea &set_target_angle_range_left(double target_angle_range_left)

Set left point value of Target Angle Range (Degree) with std::vector<double> value.

The angle range target limited to

See also

get_target_angle_range_left()

Parameters:

target_angle_range_left – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationLayOutArea& the reference of this object.

AssemblyVerificationLayOutArea &set_target_angle_range_right(double target_angle_range_right)

Set the right point value of Target Angle Range (Degree) with std::vector<double> value.

The angle range target limited to

See also

get_target_angle_range_right()

Parameters:

target_angle_range_right – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationLayOutArea& the reference of this object.

class AssemblyVerificationTemplates : public visionflow::param::SchemableParameter

AssemblyVerificationTemplates Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const TemplateMode &get_template_mode() const

Get Template Mode.

Set the Value Type of the central coordinates of the region and the width and height of the area. When the template mode is set to kValueByPixel, X-Value Y-Value W-Value H-Value should be set using pixel value. When the template mode is set to kValueByRatio, X-Value Y-Value W-Value H-Value should be set using ratio value. When the template mode is modified, X-Value Y-Value W-Value H-Value needs to be modified.

See also

set_template_mode()

Returns:

const TemplateMode & Template Mode

AssemblyVerificationTemplates &set_template_mode(const TemplateMode &template_mode)

Set Template Mode with TemplateMode value.

Set the Value Type of the central coordinates of the region and the width and height of the area. When the template mode is set to kValueByPixel, X-Value Y-Value W-Value H-Value should be set using pixel value. When the template mode is set to kValueByRatio, X-Value Y-Value W-Value H-Value should be set using ratio value. When the template mode is modified, X-Value Y-Value W-Value H-Value needs to be modified.

See also

get_template_mode()

Parameters:

template_mode – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationTemplates& the reference of this object.

const std::map<std::string, AssemblyVerificationLayOutArea> &get_areas() const

Get Areas.

The areas inside the template. The key of the map is the name of area. The value of the map is the parameter of area.

See also

set_areas()

Returns:

const std::map<std::string, AssemblyVerificationLayOutArea> & Areas

AssemblyVerificationTemplates &set_areas(const std::map<std::string, AssemblyVerificationLayOutArea> &areas)

Set Areas with std::map<std::string, AssemblyVerificationLayOutArea> value.

The areas inside the template. The key of the map is the name of area. The value of the map is the parameter of area.

See also

get_areas()

Parameters:

areas – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationTemplates& the reference of this object.

const AssemblyVerificationLayOutArea &get_areas(const std::string &key) const

Get value in Areas with key.

Warning

The key must be exist in Areas. otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const std::map<std::string, AssemblyVerificationLayOutArea> & value in Areas at key.

AssemblyVerificationTemplates &set_areas(const std::string &key, const AssemblyVerificationLayOutArea &value)

Set value in Areas with key.

Parameters:

• key – the key to set.

• value – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationTemplates& the reference of this object.

bool areas_contains(const std::string &key) const

Check if the key is exist in Areas.

Returns:

bool true if the key is exist in Areas, otherwise false.

size_t get_areas_size() const

Get the size of Areas.

Returns:

size_t the size of Areas

AssemblyVerificationLayOutArea &get_areas(const std::string &key)

Get mutable reference of value in Areas with key. A new key and default value will be created if the key does not exist.

Returns:

AssemblyVerificationLayOutArea& the mutable reference of value in Areas at key.

class AssemblyVerificationTrainingParameters : public visionflow::param::SchemableParameter

AssemblyVerificationTrainingParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int get_epoch() const

Get Epoch.

Number of epochs to train

See also

set_epoch()

Returns:

int Epoch

AssemblyVerificationTrainingParameters &set_epoch(int epoch)

Set Epoch with int value.

Number of epochs to train

See also

get_epoch()

Parameters:

epoch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationTrainingParameters& the reference of this object.

int get_batch_size() const

Get Training Batch Size.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16. The classification module needs to be set larger, generally 32, 64

See also

set_batch_size()

Returns:

int Training Batch Size

AssemblyVerificationTrainingParameters &set_batch_size(int batch_size)

Set Training Batch Size with int value.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16. The classification module needs to be set larger, generally 32, 64

See also

get_batch_size()

Parameters:

batch_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationTrainingParameters& the reference of this object.

const DataAugmentation &get_augmentations() const

Get Data Augmentation.

Data Augmentation Parameters

See also

set_augmentations()

Returns:

const DataAugmentation & Data Augmentation

AssemblyVerificationTrainingParameters &set_augmentations(const DataAugmentation &augmentations)

Set Data Augmentation with DataAugmentation value.

Data Augmentation Parameters

See also

get_augmentations()

Parameters:

augmentations – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationTrainingParameters& the reference of this object.

DataAugmentation &get_augmentations()

Get mutable reference of Data Augmentation.

Data Augmentation Parameters

Returns:

AssemblyVerificationTrainingParameters& the mutable reference of the group.

const AssemblyVerificationMaxInputSize &get_input_shape() const

Get Input Shape.

Input Shape Parameters

See also

set_input_shape()

Returns:

const AssemblyVerificationMaxInputSize & Input Shape

AssemblyVerificationTrainingParameters &set_input_shape(const AssemblyVerificationMaxInputSize &input_shape)

Set Input Shape with AssemblyVerificationMaxInputSize value.

Input Shape Parameters

See also

get_input_shape()

Parameters:

input_shape – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationTrainingParameters& the reference of this object.

AssemblyVerificationMaxInputSize &get_input_shape()

Get mutable reference of Input Shape.

Input Shape Parameters

Returns:

AssemblyVerificationTrainingParameters& the mutable reference of the group.

const AssemblyVerificationModelParameters &get_model_param() const

Get Model param.

AssemblyVerification model parameters

See also

set_model_param()

Returns:

const AssemblyVerificationModelParameters & Model param

AssemblyVerificationTrainingParameters &set_model_param(const AssemblyVerificationModelParameters &model_param)

Set Model param with AssemblyVerificationModelParameters value.

AssemblyVerification model parameters

See also

get_model_param()

Parameters:

model_param – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationTrainingParameters& the reference of this object.

AssemblyVerificationModelParameters &get_model_param()

Get mutable reference of Model param.

AssemblyVerification model parameters

Returns:

AssemblyVerificationTrainingParameters& the mutable reference of the group.

const AssemblyVerificationTargetParameters &get_target_feature() const

Get Target Feature.

AssemblyVerification Target Feature Parameters

See also

set_target_feature()

Returns:

const AssemblyVerificationTargetParameters & Target Feature

AssemblyVerificationTrainingParameters &set_target_feature(const AssemblyVerificationTargetParameters &target_feature)

Set Target Feature with AssemblyVerificationTargetParameters value.

AssemblyVerification Target Feature Parameters

See also

get_target_feature()

Parameters:

target_feature – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationTrainingParameters& the reference of this object.

AssemblyVerificationTargetParameters &get_target_feature()

Get mutable reference of Target Feature.

AssemblyVerification Target Feature Parameters

Returns:

AssemblyVerificationTrainingParameters& the mutable reference of the group.

class AssemblyVerificationFilterParameters : public visionflow::param::SchemableParameter

AssemblyVerificationFilterParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

double get_threshold() const

Get Confidence Threshold.

Keypoints with confidence greater than this threshold will be determined as a character target.

See also

set_threshold()

Returns:

double Confidence Threshold

AssemblyVerificationFilterParameters &set_threshold(double threshold)

Set Confidence Threshold with double value.

Keypoints with confidence greater than this threshold will be determined as a character target.

See also

get_threshold()

Parameters:

threshold – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationFilterParameters& the reference of this object.

double get_density() const

Get Search Density.

SKeep only one result within the radius of target size * density

See also

set_density()

Returns:

double Search Density

AssemblyVerificationFilterParameters &set_density(double density)

Set Search Density with double value.

SKeep only one result within the radius of target size * density

See also

get_density()

Parameters:

density – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

AssemblyVerificationFilterParameters& the reference of this object.

Classification Tool Parameters

class ClassificationInputShape : public visionflow::param::ISchemable

ClassificationInputShape Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Customize Input Size.

Enable this option to customize the model input shape. With this option enabled, the the input image will be resized into the customized shape before inference it.

See also

set_enable()

Returns:

bool Customize Input Size

ClassificationInputShape &set_enable(bool enable)

Set Customize Input Size with bool value.

Enable this option to customize the model input shape. With this option enabled, the the input image will be resized into the customized shape before inference it.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ClassificationInputShape& the reference of this object.

int get_base_input_width() const

Get Input Image Width.

See also

set_base_input_width()

Returns:

int Input Image Width

ClassificationInputShape &set_base_input_width(int base_input_width)

Set Input Image Width with int value.

See also

get_base_input_width()

Parameters:

base_input_width – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ClassificationInputShape& the reference of this object.

int get_base_input_height() const

Get Input Image Height.

See also

set_base_input_height()

Returns:

int Input Image Height

ClassificationInputShape &set_base_input_height(int base_input_height)

Set Input Image Height with int value.

See also

get_base_input_height()

Parameters:

base_input_height – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ClassificationInputShape& the reference of this object.

class ClassificationTrainingParameters : public visionflow::param::SchemableParameter

ClassificationTrainingParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int get_epoch() const

Get Epoch.

Number of epochs to train

See also

set_epoch()

Returns:

int Epoch

ClassificationTrainingParameters &set_epoch(int epoch)

Set Epoch with int value.

Number of epochs to train

See also

get_epoch()

Parameters:

epoch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ClassificationTrainingParameters& the reference of this object.

int get_batch_size() const

Get Training Batch Size.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16. The classification module needs to be set larger, generally 32, 64

See also

set_batch_size()

Returns:

int Training Batch Size

ClassificationTrainingParameters &set_batch_size(int batch_size)

Set Training Batch Size with int value.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16. The classification module needs to be set larger, generally 32, 64

See also

get_batch_size()

Parameters:

batch_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ClassificationTrainingParameters& the reference of this object.

const ClassificationInputShape &get_input_shape() const

Get Input Shape.

Input Shape Parameters

See also

set_input_shape()

Returns:

const ClassificationInputShape & Input Shape

ClassificationTrainingParameters &set_input_shape(const ClassificationInputShape &input_shape)

Set Input Shape with ClassificationInputShape value.

Input Shape Parameters

See also

get_input_shape()

Parameters:

input_shape – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ClassificationTrainingParameters& the reference of this object.

ClassificationInputShape &get_input_shape()

Get mutable reference of Input Shape.

Input Shape Parameters

Returns:

ClassificationTrainingParameters& the mutable reference of the group.

const TrainingMode &get_training_mode() const

Get Training Mode.

The ‘Normal’ mode is the default mode, which trains the model from scratch/pre-train. The ‘Incremental’ mode uses an incremental training algorithm to update the model parameters based on new data, without retraining the entire model from scratch.

See also

set_training_mode()

Returns:

const TrainingMode & Training Mode

ClassificationTrainingParameters &set_training_mode(const TrainingMode &training_mode)

Set Training Mode with TrainingMode value.

The ‘Normal’ mode is the default mode, which trains the model from scratch/pre-train. The ‘Incremental’ mode uses an incremental training algorithm to update the model parameters based on new data, without retraining the entire model from scratch.

See also

get_training_mode()

Parameters:

training_mode – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ClassificationTrainingParameters& the reference of this object.

bool get_only_check() const

Get Only Check.

Default is False. If set to True, the trainer will only check if it is possible to train in the given training mode without actually performing the training.

See also

set_only_check()

Returns:

bool Only Check

ClassificationTrainingParameters &set_only_check(bool only_check)

Set Only Check with bool value.

Default is False. If set to True, the trainer will only check if it is possible to train in the given training mode without actually performing the training.

See also

get_only_check()

Parameters:

only_check – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ClassificationTrainingParameters& the reference of this object.

const std::string &get_model_arch() const

Get Model Architecture.

The Fast-Model is suitable for most scenarios, but the classification accuracy may be reduced in scenarios with more categories (more than 10 categories) and lower differentiation between categories, in which you can use the High-Precision-Model, although this will result in a slightly longer training and inference time.

See also

set_model_arch()

Returns:

const std::string & Model Architecture

ClassificationTrainingParameters &set_model_arch(const std::string &model_arch)

Set Model Architecture with std::string value.

The Fast-Model is suitable for most scenarios, but the classification accuracy may be reduced in scenarios with more categories (more than 10 categories) and lower differentiation between categories, in which you can use the High-Precision-Model, although this will result in a slightly longer training and inference time.

See also

get_model_arch()

Parameters:

model_arch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ClassificationTrainingParameters& the reference of this object.

const DataAugmentation &get_augmentations() const

Get Data Augmentation.

Data Augmentation Parameters

See also

set_augmentations()

Returns:

const DataAugmentation & Data Augmentation

ClassificationTrainingParameters &set_augmentations(const DataAugmentation &augmentations)

Set Data Augmentation with DataAugmentation value.

Data Augmentation Parameters

See also

get_augmentations()

Parameters:

augmentations – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ClassificationTrainingParameters& the reference of this object.

DataAugmentation &get_augmentations()

Get mutable reference of Data Augmentation.

Data Augmentation Parameters

Returns:

ClassificationTrainingParameters& the mutable reference of the group.

bool get_small_target() const

Get Detect Small Target.

Enable this option to detect small target

See also

set_small_target()

Returns:

bool Detect Small Target

ClassificationTrainingParameters &set_small_target(bool small_target)

Set Detect Small Target with bool value.

Enable this option to detect small target

See also

get_small_target()

Parameters:

small_target – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ClassificationTrainingParameters& the reference of this object.

bool get_with_contrastive() const

Get With Contrastive Learning.

With this option enabled, The algorithm will try to improve the model performance by comparing the difference between sample and defective image. This mode is only applicable to the mixed images composed of a sample image and a defective image, it’s not applicable to single-image-samples or mixed images composed of different photos of the same defect.

See also

set_with_contrastive()

Returns:

bool With Contrastive Learning

ClassificationTrainingParameters &set_with_contrastive(bool with_contrastive)

Set With Contrastive Learning with bool value.

With this option enabled, The algorithm will try to improve the model performance by comparing the difference between sample and defective image. This mode is only applicable to the mixed images composed of a sample image and a defective image, it’s not applicable to single-image-samples or mixed images composed of different photos of the same defect.

See also

get_with_contrastive()

Parameters:

with_contrastive – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ClassificationTrainingParameters& the reference of this object.

class ClassificationInferParameters : public visionflow::param::SchemableParameter

ClassificationInferParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_visualize_param() const

Get Heat map visualization.

Heat map visualization

See also

set_visualize_param()

Returns:

bool Heat map visualization

ClassificationInferParameters &set_visualize_param(bool visualize_param)

Set Heat map visualization with bool value.

Heat map visualization

See also

get_visualize_param()

Parameters:

visualize_param – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ClassificationInferParameters& the reference of this object.

Detection Tool Parameters

class DetectionTrainingParameters : public visionflow::param::SchemableParameter

DetectionTrainingParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int get_epoch() const

Get Epoch.

Number of epochs to train

See also

set_epoch()

Returns:

int Epoch

DetectionTrainingParameters &set_epoch(int epoch)

Set Epoch with int value.

Number of epochs to train

See also

get_epoch()

Parameters:

epoch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionTrainingParameters& the reference of this object.

int get_batch_size() const

Get Training Batch Size.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16.

See also

set_batch_size()

Returns:

int Training Batch Size

DetectionTrainingParameters &set_batch_size(int batch_size)

Set Training Batch Size with int value.

The number of images involved in training in each iteration of network training. A suitable batch size can make full use of hardware and improve the convergence speed. Common values are 4, 8, and 16.

See also

get_batch_size()

Parameters:

batch_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionTrainingParameters& the reference of this object.

double get_lr_decay() const

Get Learning Rate Decay.

The learning rate is multiplied by this value every epoch. The larger the value, the faster the learning rate decreases

See also

set_lr_decay()

Returns:

double Learning Rate Decay

DetectionTrainingParameters &set_lr_decay(double lr_decay)

Set Learning Rate Decay with double value.

The learning rate is multiplied by this value every epoch. The larger the value, the faster the learning rate decreases

See also

get_lr_decay()

Parameters:

lr_decay – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionTrainingParameters& the reference of this object.

const std::string &get_model_arch() const

Get Model Architecture.

The 108 model only supports single-class detection and is fast with inference,the 110 model supports multi-class detection and more comprehensive reasoning

See also

set_model_arch()

Returns:

const std::string & Model Architecture

DetectionTrainingParameters &set_model_arch(const std::string &model_arch)

Set Model Architecture with std::string value.

The 108 model only supports single-class detection and is fast with inference,the 110 model supports multi-class detection and more comprehensive reasoning

See also

get_model_arch()

Parameters:

model_arch – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionTrainingParameters& the reference of this object.

const DataAugmentation &get_augmentations() const

Get Data Augmentation.

Data Augmentation Parameters

See also

set_augmentations()

Returns:

const DataAugmentation & Data Augmentation

DetectionTrainingParameters &set_augmentations(const DataAugmentation &augmentations)

Set Data Augmentation with DataAugmentation value.

Data Augmentation Parameters

See also

get_augmentations()

Parameters:

augmentations – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionTrainingParameters& the reference of this object.

DataAugmentation &get_augmentations()

Get mutable reference of Data Augmentation.

Data Augmentation Parameters

Returns:

DetectionTrainingParameters& the mutable reference of the group.

int get_max_size() const

Get Max Length.

To resize input images, the longer side is scaled down to match the maximum edgelength while the shorter side is scaled proportionally. If the calculated maximumedge length is smaller than the given parameter, the algorithm will automaticallyadjust the maximum edge length. Setting a larger maximum edge length can improve detection accuracy, but it also requires longer training and inference time and higher GPU memory usage.

See also

set_max_size()

Returns:

int Max Length

DetectionTrainingParameters &set_max_size(int max_size)

Set Max Length with int value.

To resize input images, the longer side is scaled down to match the maximum edgelength while the shorter side is scaled proportionally. If the calculated maximumedge length is smaller than the given parameter, the algorithm will automaticallyadjust the maximum edge length. Setting a larger maximum edge length can improve detection accuracy, but it also requires longer training and inference time and higher GPU memory usage.

See also

get_max_size()

Parameters:

max_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionTrainingParameters& the reference of this object.

const DetectionInputShape &get_input_shape() const

Get Input Shape.

Input Shape Parameters

See also

set_input_shape()

Returns:

const DetectionInputShape & Input Shape

DetectionTrainingParameters &set_input_shape(const DetectionInputShape &input_shape)

Set Input Shape with DetectionInputShape value.

Input Shape Parameters

See also

get_input_shape()

Parameters:

input_shape – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionTrainingParameters& the reference of this object.

DetectionInputShape &get_input_shape()

Get mutable reference of Input Shape.

Input Shape Parameters

Returns:

DetectionTrainingParameters& the mutable reference of the group.

const DetectionTrainingStrategy &get_train_tricks() const

Get train tricks.

Detection Strategy Parameters

See also

set_train_tricks()

Returns:

const DetectionTrainingStrategy & train tricks

DetectionTrainingParameters &set_train_tricks(const DetectionTrainingStrategy &train_tricks)

Set train tricks with DetectionTrainingStrategy value.

Detection Strategy Parameters

See also

get_train_tricks()

Parameters:

train_tricks – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionTrainingParameters& the reference of this object.

DetectionTrainingStrategy &get_train_tricks()

Get mutable reference of train tricks.

Detection Strategy Parameters

Returns:

DetectionTrainingParameters& the mutable reference of the group.

class DetectionInferParameters : public visionflow::param::SchemableParameter

DetectionInferParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

double get_confidence_threshold() const

Get Confidence Threshold.

Only retain objects with a confidence score greater than this threshold.

See also

set_confidence_threshold()

Returns:

double Confidence Threshold

DetectionInferParameters &set_confidence_threshold(double confidence_threshold)

Set Confidence Threshold with double value.

Only retain objects with a confidence score greater than this threshold.

See also

get_confidence_threshold()

Parameters:

confidence_threshold – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionInferParameters& the reference of this object.

int get_min_target_size() const

Get Minimum Target Size.

Only retain objects with a width and height greater than this minimum size.

See also

set_min_target_size()

Returns:

int Minimum Target Size

DetectionInferParameters &set_min_target_size(int min_target_size)

Set Minimum Target Size with int value.

Only retain objects with a width and height greater than this minimum size.

See also

get_min_target_size()

Parameters:

min_target_size – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionInferParameters& the reference of this object.

int get_max_target_num() const

Get Maximum Target Count.

Retain only the top N targets with the highest confidence score

See also

set_max_target_num()

Returns:

int Maximum Target Count

DetectionInferParameters &set_max_target_num(int max_target_num)

Set Maximum Target Count with int value.

Retain only the top N targets with the highest confidence score

See also

get_max_target_num()

Parameters:

max_target_num – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionInferParameters& the reference of this object.

class DetectionTrainingStrategy : public visionflow::param::ISchemable

DetectionTrainingStrategy Parameter class generated by jinja2 automatically.

Detection training tricks

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_auto_anchor() const

Get Auto Anchor.

Auto anchor will automatically adjust the preset anchor size based on the image label

See also

set_auto_anchor()

Returns:

bool Auto Anchor

DetectionTrainingStrategy &set_auto_anchor(bool auto_anchor)

Set Auto Anchor with bool value.

Auto anchor will automatically adjust the preset anchor size based on the image label

See also

get_auto_anchor()

Parameters:

auto_anchor – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionTrainingStrategy& the reference of this object.

class DetectionInputShape : public visionflow::param::ISchemable

DetectionInputShape Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable Custom Input Shape.

Enable to Customize Network Input Shape

See also

set_enable()

Returns:

bool Enable Custom Input Shape

DetectionInputShape &set_enable(bool enable)

Set Enable Custom Input Shape with bool value.

Enable to Customize Network Input Shape

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionInputShape& the reference of this object.

int get_base_input_width() const

Get Base Input Width.

Base Input Width

See also

set_base_input_width()

Returns:

int Base Input Width

DetectionInputShape &set_base_input_width(int base_input_width)

Set Base Input Width with int value.

Base Input Width

See also

get_base_input_width()

Parameters:

base_input_width – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionInputShape& the reference of this object.

int get_base_input_height() const

Get Base Input Height.

Base Input Height

See also

set_base_input_height()

Returns:

int Base Input Height

DetectionInputShape &set_base_input_height(int base_input_height)

Set Base Input Height with int value.

Base Input Height

See also

get_base_input_height()

Parameters:

base_input_height – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

DetectionInputShape& the reference of this object.

View Transformer Parameters

enum visionflow::param::ViewTransMode

Values:

enumerator kByPixel = 0

enumerator kByRatio = 1

class ViewFilterParameters : public visionflow::param::SchemableParameter

ViewFilterParameters Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_with_ok_view() const

Get With OK View.

Treating views with no detection results as detection targets

See also

set_with_ok_view()

Returns:

bool With OK View

ViewFilterParameters &set_with_ok_view(bool with_ok_view)

Set With OK View with bool value.

Treating views with no detection results as detection targets

See also

get_with_ok_view()

Parameters:

with_ok_view – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ViewFilterParameters& the reference of this object.

const std::map<std::string, SingleClassPolygonsFilterParameters> &get_keep_classes() const

Get Keep Classes.

Polygon thresholds (area, side-length,etc.) named with the label classes according to which to filter the polygons.

See also

set_keep_classes()

Returns:

const std::map<std::string, SingleClassPolygonsFilterParameters> & Keep Classes

ViewFilterParameters &set_keep_classes(const std::map<std::string, SingleClassPolygonsFilterParameters> &keep_classes)

Set Keep Classes with std::map<std::string, SingleClassPolygonsFilterParameters> value.

Polygon thresholds (area, side-length,etc.) named with the label classes according to which to filter the polygons.

See also

get_keep_classes()

Parameters:

keep_classes – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ViewFilterParameters& the reference of this object.

const SingleClassPolygonsFilterParameters &get_keep_classes(const std::string &key) const

Get value in Keep Classes with key.

Warning

The key must be exist in Keep Classes. otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const std::map<std::string, SingleClassPolygonsFilterParameters> & value in Keep Classes at key.

ViewFilterParameters &set_keep_classes(const std::string &key, const SingleClassPolygonsFilterParameters &value)

Set value in Keep Classes with key.

Parameters:

• key – the key to set.

• value – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ViewFilterParameters& the reference of this object.

bool keep_classes_contains(const std::string &key) const

Check if the key is exist in Keep Classes.

Returns:

bool true if the key is exist in Keep Classes, otherwise false.

size_t get_keep_classes_size() const

Get the size of Keep Classes.

Returns:

size_t the size of Keep Classes

SingleClassPolygonsFilterParameters &get_keep_classes(const std::string &key)

Get mutable reference of value in Keep Classes with key. A new key and default value will be created if the key does not exist.

Returns:

SingleClassPolygonsFilterParameters& the mutable reference of value in Keep Classes at key.

class ViewTransAutoMask

ViewTransAutoMask Parameter class generated by jinja2 automatically.

Public Functions

bool get_enable() const

Get Anable View Auto Mask.

Enable this option to use the detection result from the previous module as a mask.

See also

set_enable()

Returns:

bool Anable View Auto Mask

ViewTransAutoMask &set_enable(bool enable)

Set Anable View Auto Mask with bool value.

Enable this option to use the detection result from the previous module as a mask.

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ViewTransAutoMask& the reference of this object.

int get_dilate_pixel() const

Get Dilate Pixel.

Dilate the detection result from the previous module before mask. A negative value indicate erosion operation.

See also

set_dilate_pixel()

Returns:

int Dilate Pixel

ViewTransAutoMask &set_dilate_pixel(int dilate_pixel)

Set Dilate Pixel with int value.

Dilate the detection result from the previous module before mask. A negative value indicate erosion operation.

See also

get_dilate_pixel()

Parameters:

dilate_pixel – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ViewTransAutoMask& the reference of this object.

bool get_reverse_mask() const

Get Reverse Mask.

Inverse masking of areas in the view that have no detection results from the previous module. Note that this option works after the dilate option if it’s not zero.

See also

set_reverse_mask()

Returns:

bool Reverse Mask

ViewTransAutoMask &set_reverse_mask(bool reverse_mask)

Set Reverse Mask with bool value.

Inverse masking of areas in the view that have no detection results from the previous module. Note that this option works after the dilate option if it’s not zero.

See also

get_reverse_mask()

Parameters:

reverse_mask – the value to set.

Returns:

ViewTransAutoMask& the reference of this object.

class ViewTransformParameter

ViewTransformParameter Parameter class generated by jinja2 automatically.

Public Functions

ViewTransMode get_mode() const

Get Offset Mode.

Offset by a fixed number of pixels or by a percentage depending on the size of the defect.

See also

set_mode()

Returns:

ViewTransMode View Mode

ViewTransformParameter &set_mode(ViewTransMode mode)

Set the view transform mode as ByPixel or ByRatio.

Generate view by a fixed number of pixels or by a percentage depending on the size of the defect.

See also

get_mode()

Parameters:

mode – the The View mode.

Returns:

ViewTransformParameter& the reference of this object.

const geometry::Vector2f &get_offset() const

Get Offset Vector.

See also

set_offset()

Returns:

The offset vector, Its practical use is determined by the VeiwTransOffsetMode in get_mode()

ViewTransformParameter &set_offset(const geometry::Vector2f &offset)

Set View offset vector With the centre of the original test result as the reference coordinate system.

See also

get_offset()

Parameters:

offset – the offset value to set.

Returns:

ViewTransformParameter& the reference of this object.

const geometry::Size2f &get_window_size() const

Get Window Size.

The view window size, The meaning of the options is determined by the Window Size Mode

The View window size mode. “DynamicPixel” mode means get the window size from the detection result from previous module and you can scale it with a fixed number of pixels; “DynamicRatio” mode means get the window size from the detection result from previous module and you can scale it by ratio; “FixedWindow” means set the view windows to a fixed size regardless of the size of the detection result from previous module.

See also

set_window_size()

Returns:

Window Size or Window expand size base on the detection result.

ViewTransformParameter &set_window_size(const geometry::Size2f &window_size)

Set Window size.

The view window size, The meaning of the options is determined by the view mode.

See also

get_window_size()

Parameters:

window_size – the value to set.

Returns:

ViewTransformParameter& the reference of this object.

bool get_fixed_window_size() const

Get If the view windows size is fixed.

Returns:

bool

ViewTransformParameter &set_fixed_window_size(bool fixed_window_size)

Set this option as true to fix the view window size.

Parameters:

fixed_window_size –

Returns:

ViewTransformParameter&

geometry::Radian get_rotate_angle() const

Get the view rotate angle.

Returns:

geometry::Radian the view rotate angle base on the reference coordinate system.

ViewTransformParameter &set_rotate_angle(const geometry::Radian &angle)

Set the view rotate angle.

Parameters:

angle – rotate angle.

Returns:

ViewTransformParameter& this parameter object.

const ViewTransAutoMask &get_auto_mask() const

Get Auto View Mask.

Auto view mask parameters

See also

set_auto_mask()

Returns:

const ViewTransAutoMask & Auto View Mask

ViewTransformParameter &set_auto_mask(const ViewTransAutoMask &auto_mask)

Set Auto View Mask with ViewTransAutoMask value.

Auto view mask parameters

See also

get_auto_mask()

Parameters:

auto_mask – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ViewTransformParameter& the reference of this object.

ViewTransAutoMask &get_auto_mask()

Get mutable reference of Auto View Mask.

Auto view mask parameters

Returns:

ViewTransformParameter& the mutable reference of the group.

const geometry::MultiPolygon2f &get_manual_mask() const

Get the manual mask object.

Returns:

const geometry::MultiPolygon2f& Get the manual view mask.

ViewTransformParameter &set_manual_mask(const geometry::MultiPolygon2f &manual_mask)

Set the manual mask object. The x-y coordinate value of the points in MultiPolygon2f should be established in the coordinate system with the upper-left corner point of the view after transforming based on the defact region and its angle.

class ViewTransformParameterList : public visionflow::param::IParameter

Public Functions

size_t size() const

View transform parameter size.

Returns:

size_t

std::vector<std::string> keys() const

All view transform parameters’ names.

Returns:

std::vector<std::string> parameter names.

const ViewTransformParameter &at(const std::string &view_param_name) const

Get view specified by the name.

Parameters:

view_param_name –

Returns:

ViewTransformParameter const&

bool contains(const std::string &view_param_name) const

Check if the list contains parameter specified by the name.

Parameters:

view_param_name –

Returns:

bool

void erase(const std::string &view_param_name)

remove parameter from the list.

std::string add(const ViewTransformParameter &view_param)

Add a new ViewTransformParameter instance into this list. An new id will be generated for the parameter added.

Parameters:

view_param – The parameter to be added into this list.

Returns:

std::string The parameter id generated by the list.

void update(const std::string &view_param_name, const ViewTransformParameter &view_param)

Update view parameter specified by view_param_name with the new parameter. If parameter specified by view_param_name does not exist, the new ViewTransformParameter instance will be added into this list with key view_param_name.

Parameters:

• view_param_name – The key of the parameter to be updated.

• view_param – The new parameter to be updated.

Sample Recommendation Parameters:

class SampleRecommendationParameter : public visionflow::param::SchemableParameter

SampleRecommendationParameter Parameter class generated by jinja2 automatically.

Parameter for the sample recommendation algorithm

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

double get_recommend_budget() const

Get Recommend Budget.

The ratio of views to recommend

See also

set_recommend_budget()

Returns:

double Recommend Budget

SampleRecommendationParameter &set_recommend_budget(double recommend_budget)

Set Recommend Budget with double value.

The ratio of views to recommend

See also

get_recommend_budget()

Parameters:

recommend_budget – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

SampleRecommendationParameter& the reference of this object.

class PropertyObjectId : public visionflow::param::ISchemable

PropertyObjectId Parameter class generated by jinja2 automatically.

Parameter group for a object IDs in a property.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int get_property_id() const

Get Property ID.

ID of the property

See also

set_property_id()

Returns:

int Property ID

PropertyObjectId &set_property_id(int property_id)

Set Property ID with int value.

ID of the property

See also

get_property_id()

Parameters:

property_id – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

PropertyObjectId& the reference of this object.

const std::vector<std::string> &get_object_ids() const

Get Object IDs.

IDs of the objects

See also

set_object_ids()

Returns:

const std::vector<std::string> & Object IDs

PropertyObjectId &set_object_ids(const std::vector<std::string> &object_ids)

Set Object IDs with std::vector<std::string> value.

IDs of the objects

See also

get_object_ids()

Parameters:

object_ids – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

PropertyObjectId& the reference of this object.

const std::string &get_object_ids(size_t index) const

Get value in Object IDs with index.

Warning

The index must be less than get_object_ids_size(). otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const std::string & value in Object IDs at index.

size_t get_object_ids_size() const

Get the size of Object IDs.

Returns:

size_t the size of Object IDs

class PropertyObjectIdSet : public visionflow::param::SchemableParameter

PropertyObjectIdSet Parameter class generated by jinja2 automatically.

Parameter group for a set of property object IDs.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const std::vector<PropertyObjectId> &get_object_id_sets() const

Get Property Object ID Sets.

Sets of object IDs in properties

See also

set_object_id_sets()

Returns:

const std::vector<PropertyObjectId> & Property Object ID Sets

PropertyObjectIdSet &set_object_id_sets(const std::vector<PropertyObjectId> &object_id_sets)

Set Property Object ID Sets with std::vector<PropertyObjectId> value.

Sets of object IDs in properties

See also

get_object_id_sets()

Parameters:

object_id_sets – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

PropertyObjectIdSet& the reference of this object.

const PropertyObjectId &get_object_id_sets(size_t index) const

Get value in Property Object ID Sets with index.

Warning

The index must be less than get_object_id_sets_size(). otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

const PropertyObjectId & value in Property Object ID Sets at index.

size_t get_object_id_sets_size() const

Get the size of Property Object ID Sets.

Returns:

size_t the size of Property Object ID Sets

PropertyObjectId &get_object_id_sets(size_t index)

Get mutable reference of value in Property Object ID Sets with index.

Warning

The index must be less than get_object_id_sets_size(). otherwise, the behavior is undefined and will cause segmentation fault.

Returns:

PropertyObjectId& the mutable reference of value in Property Object ID Sets at index.

Training Set Recommend Parameters:

enum visionflow::param::TrainingSetRecommendType

Values:

enumerator kViewBase = 0

class TrainingSetRecommendParameter : public visionflow::param::SchemableParameter

TrainingSetRecommendParameter Parameter class generated by jinja2 automatically.

Parameter for the auto class balance

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

int get_num_to_select() const

Get Select Number.

The number of views to select

See also

set_num_to_select()

Returns:

int Select Number

TrainingSetRecommendParameter &set_num_to_select(int num_to_select)

Set Select Number with int value.

The number of views to select

See also

get_num_to_select()

Parameters:

num_to_select – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

TrainingSetRecommendParameter& the reference of this object.

const TrainingSetRecommendType &get_algorithm_type() const

Get Algorithm type.

The Algorithm type string to be used training set recommend, currently supports three types: ViewBase, PixelBase, and RegionBase.

See also

set_algorithm_type()

Returns:

const TrainingSetRecommendType & Algorithm type

TrainingSetRecommendParameter &set_algorithm_type(const TrainingSetRecommendType &algorithm_type)

Set Algorithm type with TrainingSetRecommendType value.

The Algorithm type string to be used training set recommend, currently supports three types: ViewBase, PixelBase, and RegionBase.

See also

get_algorithm_type()

Parameters:

algorithm_type – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

TrainingSetRecommendParameter& the reference of this object.

Integration Tool Parameters:

class IntegrationClassifyParameter : public visionflow::param::SchemableParameter

IntegrationClassifyParameter Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

const std::string &get_script() const

 @brief Get Category Script.

 The list of classes and conditions to be evaluated. Note that the function `category_sample` must exist and return a string value.

And we provide a sample script to help you to implement your own script:

import visionflow as vf
from visionflow import geometry as geo

def region_covered_by(region, ring):
    output = geo.MultiPolygon2f()
    geo.difference(ring, region.polygon(), output)
    return len(output) == 0

classifier_condition = {
    "Ok": lambda sample: any(
        0 < region.area() < 10000
        and 0 < region.polygon().bounding_box().long_side() < 30000
        and 0 < region.polygon().bounding_box().short_side() < 30000
        and -45 < region.angle().degree() < 45
        and 0.1 < region.score() < 1.0
        and region_covered_by(region, geo.Ring2f([geo.Point2f(0,0), geo.Point2f(0,100), geo.Point2f(100,100), geo.Point2f(100,0)]))
        for id, region in sample["Segmentation/pred"]
    ),
    "Car": lambda sample: any(
        0 < region.area() < 10000
        and 0 < region.polygon().bounding_box().long_side() < 30000
        and 0 < region.polygon().bounding_box().short_side() < 30000
        and -45 < region.angle().degree() < 45
        and 0.1 < region.score() < 1.0
        and region_covered_by(region, geo.Ring2f([geo.Point2f(0,0), geo.Point2f(0,200), geo.Point2f(200,200), geo.Point2f(200,0)]))
        for id, region in sample["Segmentation/pred"]
    )
    or all(
        0 < region.area() < 10000
        and 0 < region.polygon().bounding_box().long_side() < 30000
        and 0 < region.polygon().bounding_box().short_side() < 30000
        and -45 < region.angle().degree() < 45
        and 0.1 < region.score() < 1.0
        and region_covered_by(region, geo.Ring2f([geo.Point2f(0,0), geo.Point2f(0,200), geo.Point2f(200,200), geo.Point2f(200,0)]))
        for id, region in sample["Classification/pred"]
    ),
}


# 综合判定节点的入口函数，不要修改这个函数的名称、参数和返回值类型
def integration_classifier(sample: vf.ISample) -> str:
    for category, condition in classifier_condition.items():
        if condition(sample):
            return category
    return "Other"

See also

set_script()

Returns:

const std::string & Category Script

IntegrationClassifyParameter &set_script(const std::string &script)

 @brief Set Category Script with std::string value.

 The list of classes and conditions to be evaluated. Note that the function `category_sample` must exist and return a string value.

And we provide a sample script to help you to implement your own script:

import visionflow as vf
from visionflow import geometry as geo

def region_covered_by(region, ring):
    output = geo.MultiPolygon2f()
    geo.difference(ring, region.polygon(), output)
    return len(output) == 0

classifier_condition = {
    "Ok": lambda sample: any(
        0 < region.area() < 10000
        and 0 < region.polygon().bounding_box().long_side() < 30000
        and 0 < region.polygon().bounding_box().short_side() < 30000
        and -45 < region.angle().degree() < 45
        and 0.1 < region.score() < 1.0
        and region_covered_by(region, geo.Ring2f([geo.Point2f(0,0), geo.Point2f(0,100), geo.Point2f(100,100), geo.Point2f(100,0)]))
        for id, region in sample["Segmentation/pred"]
    ),
    "Car": lambda sample: any(
        0 < region.area() < 10000
        and 0 < region.polygon().bounding_box().long_side() < 30000
        and 0 < region.polygon().bounding_box().short_side() < 30000
        and -45 < region.angle().degree() < 45
        and 0.1 < region.score() < 1.0
        and region_covered_by(region, geo.Ring2f([geo.Point2f(0,0), geo.Point2f(0,200), geo.Point2f(200,200), geo.Point2f(200,0)]))
        for id, region in sample["Segmentation/pred"]
    )
    or all(
        0 < region.area() < 10000
        and 0 < region.polygon().bounding_box().long_side() < 30000
        and 0 < region.polygon().bounding_box().short_side() < 30000
        and -45 < region.angle().degree() < 45
        and 0.1 < region.score() < 1.0
        and region_covered_by(region, geo.Ring2f([geo.Point2f(0,0), geo.Point2f(0,200), geo.Point2f(200,200), geo.Point2f(200,0)]))
        for id, region in sample["Classification/pred"]
    ),
}


# 综合判定节点的入口函数，不要修改这个函数的名称、参数和返回值类型
def integration_classifier(sample: vf.ISample) -> str:
    for category, condition in classifier_condition.items():
        if condition(sample):
            return category
    return "Other"

See also

get_script()

Parameters:

script – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

IntegrationClassifyParameter& the reference of this object.

const std::string &get_unmatched_name() const

Get Unmatched Class Name.

The class name string to be used if none of the conditions are met.

See also

set_unmatched_name()

Returns:

const std::string & Unmatched Class Name

IntegrationClassifyParameter &set_unmatched_name(const std::string &unmatched_name)

Set Unmatched Class Name with std::string value.

The class name string to be used if none of the conditions are met.

See also

get_unmatched_name()

Parameters:

unmatched_name – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

IntegrationClassifyParameter& the reference of this object.

以下参数项仅为示例

enum visionflow::param::ExampleAugmentType

Values:

enumerator kExampleSelect = 0

enumerator kExampleSecondSelect = 1

enumerator kExampleOther = 2

class ExampleAugRotate : public visionflow::param::ISchemable

ExampleAugRotate Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable.

Enable this augments

See also

set_enable()

Returns:

bool Enable

ExampleAugRotate &set_enable(bool enable)

Set Enable with bool value.

Enable this augments

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ExampleAugRotate& the reference of this object.

double get_rotate_angle() const

Get Rotate Angle.

Rotate angle

See also

set_rotate_angle()

Returns:

double Rotate Angle

ExampleAugRotate &set_rotate_angle(double rotate_angle)

Set Rotate Angle with double value.

Rotate angle

See also

get_rotate_angle()

Parameters:

rotate_angle – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ExampleAugRotate& the reference of this object.

bool get_flip_horizontal() const

Get Flip Horizontal.

Flip horizontal

See also

set_flip_horizontal()

Returns:

bool Flip Horizontal

ExampleAugRotate &set_flip_horizontal(bool flip_horizontal)

Set Flip Horizontal with bool value.

Flip horizontal

See also

get_flip_horizontal()

Parameters:

flip_horizontal – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ExampleAugRotate& the reference of this object.

bool get_flip_vertical() const

Get Flip Vertical.

Flip vertical

See also

set_flip_vertical()

Returns:

bool Flip Vertical

ExampleAugRotate &set_flip_vertical(bool flip_vertical)

Set Flip Vertical with bool value.

Flip vertical

See also

get_flip_vertical()

Parameters:

flip_vertical – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ExampleAugRotate& the reference of this object.

class ExampleAugShift : public visionflow::param::ISchemable

ExampleAugShift Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable.

Enable this augments

See also

set_enable()

Returns:

bool Enable

ExampleAugShift &set_enable(bool enable)

Set Enable with bool value.

Enable this augments

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ExampleAugShift& the reference of this object.

double get_shift_width() const

Get Shift Width.

Shift width

See also

set_shift_width()

Returns:

double Shift Width

ExampleAugShift &set_shift_width(double shift_width)

Set Shift Width with double value.

Shift width

See also

get_shift_width()

Parameters:

shift_width – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ExampleAugShift& the reference of this object.

double get_shift_height() const

Get Shift Height.

Shift height

See also

set_shift_height()

Returns:

double Shift Height

ExampleAugShift &set_shift_height(double shift_height)

Set Shift Height with double value.

Shift height

See also

get_shift_height()

Parameters:

shift_height – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ExampleAugShift& the reference of this object.

class ExampleAugments : public visionflow::param::SchemableParameter

ExampleAugments Parameter class generated by jinja2 automatically.

Public Functions

virtual const std::string &schema_str() const override

Get the json schema of the class.

Returns:

std::string the json schema string.

virtual const void *schema() const override

Get nlohmann::json object pointer.

As most user not config the nlohmann::json library, we convert it as void pointer to avoid exposing types from third-party libraries to users.

Note

As this interface is a virtual function, we can not hide it with macro.

Warning

Do not edit return value.

Returns:

void* The nlohmann::json object pointer

bool get_enable() const

Get Enable.

Enable this augments

See also

set_enable()

Returns:

bool Enable

ExampleAugments &set_enable(bool enable)

Set Enable with bool value.

Enable this augments

See also

get_enable()

Parameters:

enable – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ExampleAugments& the reference of this object.

const ExampleAugmentType &get_example_select() const

Get Example Select Option.

Example select option

See also

set_example_select()

Returns:

const ExampleAugmentType & Example Select Option

ExampleAugments &set_example_select(const ExampleAugmentType &example_select)

Set Example Select Option with ExampleAugmentType value.

Example select option

See also

get_example_select()

Parameters:

example_select – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ExampleAugments& the reference of this object.

const ExampleAugShift &get_shift() const

Get Shift.

Shift augments

See also

set_shift()

Returns:

const ExampleAugShift & Shift

ExampleAugments &set_shift(const ExampleAugShift &shift)

Set Shift with ExampleAugShift value.

Shift augments

See also

get_shift()

Parameters:

shift – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ExampleAugments& the reference of this object.

ExampleAugShift &get_shift()

Get mutable reference of Shift.

Shift augments

Returns:

ExampleAugments& the mutable reference of the group.

const ExampleAugRotate &get_ctm_rotate() const

Get CustomizeRotate.

Rotate augments

See also

set_ctm_rotate()

Returns:

const ExampleAugRotate & CustomizeRotate

ExampleAugments &set_ctm_rotate(const ExampleAugRotate &ctm_rotate)

Set CustomizeRotate with ExampleAugRotate value.

Rotate augments

See also

get_ctm_rotate()

Parameters:

ctm_rotate – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ExampleAugments& the reference of this object.

ExampleAugRotate &get_ctm_rotate()

Get mutable reference of CustomizeRotate.

Rotate augments

Returns:

ExampleAugments& the mutable reference of the group.

const ExampleAugRotate &get_rotate() const

Get Rotate.

Rotate augments

See also

set_rotate()

Returns:

const ExampleAugRotate & Rotate

ExampleAugments &set_rotate(const ExampleAugRotate &rotate)

Set Rotate with ExampleAugRotate value.

Rotate augments

See also

get_rotate()

Parameters:

rotate – the value to set.

Throws:

visionflow::excepts::JsonSchemaValidationError – if validate the value with json schema failed.

Returns:

ExampleAugments& the reference of this object.

ExampleAugRotate &get_rotate()

Get mutable reference of Rotate.

Rotate augments

Returns:

ExampleAugments& the mutable reference of the group.