A ETextReader aims to find and read texts inside a given ROI, through ETextReader::Read. The returned texts can be filtered by providing a ETopology. An optimization phase can be performed before starting to read, to increase the processing speed of ETextReader. This optimization phase can be launched with ETextReader::Optimize. More...

#include <eVision/EasyOcr/EasyDeepOCR/ETextReader.h>

Collaboration diagram for eVision::EasyDeepOCR::ETextReader:

Public Member Functions
	ETextReader ()
	Creates a ETextReader.
	ETextReader (const eVision::EasyDeepOCR::ETextReader &other)
	Creates a ETextReader.
virtual	~ETextReader () throw ()
eVision::EasyDeepOCR::ETextReader &	operator= (const eVision::EasyDeepOCR::ETextReader &other)
	Assignment operator.
bool	operator== (const eVision::EasyDeepOCR::ETextReader &other) const
	Checks if this ETextReader instance is strictly equal to another ETextReader instance.
bool	IsEqual (const eVision::EasyDeepOCR::ETextReader &other) const
	Checks if this ETextReader instance is strictly equal to another ETextReader instance.
void	Load (const std::string &path)
	Loads the ETextReader.
void	Load (eVision::ESerializer *serializer)
	Loads the ETextReader.
void	Save (const std::string &path)
	Saves the ETextReader.
void	Save (eVision::ESerializer *serializer)
	Saves the ETextReader.
void	Initialize ()
	Initializes the ETextReader. ETextReader must be initialized before using ETextReader::Read.
bool	IsInitialized () const
	Returns true if the ETextReader is initialized.
void	SetTopology (const eVision::ETopology &topology)
	ETopology of the ETextReader.
eVision::ETopology	GetTopology () const
	ETopology of the ETextReader.
bool	GetHasTopology () const
	Returns true if the ETopology is set.
void	ResetTopology ()
	Forgets the ETopology previously set.
void	SetExecutionSettings (const eVision::EasyDeepLearning::EDeepLearningExecutionSettings &executionSettings)
	EDeepLearningExecutionSettings used by the ETextReader.
eVision::EasyDeepLearning::EDeepLearningExecutionSettings	GetExecutionSettings () const
	EDeepLearningExecutionSettings used by the ETextReader.
void	SetTextLineHeight (int minHeight, int maxHeight)
	Sets the range for the text line height. All returned text lines will have a height in the range [minHeight, maxHeight]. Any text line that does not fit this constraint, will be discarded. By default, it is not set, so no filtering is applied.
eVision::EFloatRange	GetTextLineHeight () const
	Gets the range for the text line height. Throws if not previously set. By default, it is not set, so no filtering is applied.
void	SetTextAngle (float arg0)
	Angle hint for the text to be found.
float	GetTextAngle () const
	Angle hint for the text to be found.
bool	GetHasTextAngle () const
	Returns true if the text angle hint is defined.
void	UnsetTextAngle ()
	Unsets the text angle hint.
eVision::EasyDeepOCR::EOptimizationReport	Optimize (const eVision::EasyDeepOCR::ESampleList &samples)
	Sets the optimal internal parameters by finding the text in the corresponding input image, for each ESample.
eVision::EasyDeepOCR::EOptimizationReport	Optimize (const eVision::EasyDeepOCR::ESampleList &samples, eVision::EImageType imageType)
	Sets the optimal internal parameters by finding the text in the corresponding input image, for each ESample.
eVision::EasyDeepOCR::EOptimizationReport	Optimize (const eVision::EasyDeepOCR::EOptimizationReport &optimizationReport)
	Sets the optimal internal parameters by finding the text in the corresponding input image, for each ESample.
bool	IsOptimized () const
	Returns true if the ETextReader has been optimized with a call to ETextReader::Optimize, and false otherwise.
void	ResetOptimization ()
	Forgets the optimal parameters previously set by ETextReader::Optimize.
void	SetLocationCustomModelPath (const std::string &modelPath)
	Sets a location custom model using the path to a saved EDeepOCRTool (user trained model).
void	SetLocationCustomModel (const eVision::EasyDeepLearning::EDeepOCRTool &tool)
	Sets a location custom model from a EDeepOCRTool object (user trained model).
bool	GetHasLocationCustomModel () const
	Whether the reader has a location custom model.
bool	HasLocationCustomModel () const
	Whether the reader has a location custom model.
void	ResetLocationCustomModel ()
	Removes any location custom model that was previously set and use the default location model.
void	SetRecognitionCustomModelPath (const std::string &modelPath)
	Sets a recognition custom model using the path to a saved EDeepOCRTool (user trained model).
void	SetRecognitionCustomModel (const eVision::EasyDeepLearning::EDeepOCRTool &tool)
	Sets a recognition custom model from a EDeepOCRTool object (user trained model).
bool	GetHasRecognitionCustomModel () const
	Whether the reader has a recognition custom model.
bool	HasRecognitionCustomModel () const
	Whether the reader has a recognition custom model.
void	ResetRecognitionCustomModel ()
	Removes any recognition custom model that was previously set and use the default recognition model.
void	SetOrientationCustomModelPath (const std::string &modelPath)
	Sets an orientation custom model using the path to a saved EDeepOCRTool (user trained model).
void	SetOrientationCustomModel (const eVision::EasyDeepLearning::EDeepOCRTool &tool)
	Sets an orientation custom model from a EDeepOCRTool object (user trained model).
bool	GetHasOrientationCustomModel () const
	Whether the reader has a orientation custom model.
bool	HasOrientationCustomModel () const
	Whether the reader has a orientation custom model.
void	ResetOrientationCustomModel ()
	Removes any orientation custom model that was previously set and use the default orientation model.
std::vector< eVision::EasyDeepOCR::EText >	Read (const eVision::EBaseROI &roi) const
	Reads texts inside an EBaseROI. Texts are returned as EText. Texts comply with the ETopology previously specified with ETextReader::SetTopology, if any. The processing speed can be increased with ETextReader::Optimize.
std::vector< eVision::EasyDeepOCR::EText >	Read (const eVision::EBaseROI &roi, const eVision::ERegion &region) const
	Reads texts inside an EBaseROI. Texts are returned as EText. Texts comply with the ETopology previously specified with ETextReader::SetTopology, if any. The processing speed can be increased with ETextReader::Optimize.
std::vector< eVision::EasyDeepOCR::EText >	Read (const eVision::EBaseROI &roi, const eVision::ERegion &region, const eVision::EFrame &frame) const
	Reads texts inside an EBaseROI. Texts are returned as EText. Texts comply with the ETopology previously specified with ETextReader::SetTopology, if any. The processing speed can be increased with ETextReader::Optimize.
std::vector< eVision::EasyDeepOCR::EText >	Read (const eVision::EBaseROI &roi, const eVision::EFrame &frame) const
	Reads texts inside an EBaseROI. Texts are returned as EText. Texts comply with the ETopology previously specified with ETextReader::SetTopology, if any. The processing speed can be increased with ETextReader::Optimize.
std::vector< eVision::ERectangle >	Localize (const eVision::EBaseROI &roi) const
	Localizes the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.
std::vector< eVision::ERectangle >	Localize (const eVision::EBaseROI &roi, const eVision::ERegion &region) const
	Localizes the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.
std::vector< eVision::ERectangle >	Localize (const eVision::EBaseROI &roi, const eVision::ERegion &region, const eVision::EFrame &frame) const
	Localizes the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.
std::vector< eVision::ERectangle >	Localize (const eVision::EBaseROI &roi, const eVision::EFrame &frame) const
	Localizes the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.
std::vector< eVision::ERectangle >	Locate (const eVision::EBaseROI &roi) const
	Locates the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.
std::vector< eVision::ERectangle >	Locate (const eVision::EBaseROI &roi, const eVision::ERegion &region) const
	Locates the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.
std::vector< eVision::ERectangle >	Locate (const eVision::EBaseROI &roi, const eVision::ERegion &region, const eVision::EFrame &frame) const
	Locates the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.
std::vector< eVision::ERectangle >	Locate (const eVision::EBaseROI &roi, const eVision::EFrame &frame) const
	Locates the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &miniature, bool automaticReadingDirection) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &miniature) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &miniature, uint32_t topologyLineIndex, bool automaticReadingDirection) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &miniature, uint32_t topologyLineIndex) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERectangle &rectangle, bool automaticReadingDirection) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERectangle &rectangle) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERectangle &rectangle, uint32_t topologyLineIndex, bool automaticReadingDirection) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERectangle &rectangle, uint32_t topologyLineIndex) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERectangle &rectangle, const eVision::EFrame &frame, bool automaticReadingDirection) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERectangle &rectangle, const eVision::EFrame &frame) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERectangle &rectangle, const eVision::EFrame &frame, uint32_t topologyLineIndex, bool automaticReadingDirection) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERectangle &rectangle, const eVision::EFrame &frame, uint32_t topologyLineIndex) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERegion &region, bool automaticReadingDirection) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERegion &region) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERegion &region, uint32_t topologyLineIndex, bool automaticReadingDirection) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERegion &region, uint32_t topologyLineIndex) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERegion &region, const eVision::EFrame &frame, bool automaticReadingDirection) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERegion &region, const eVision::EFrame &frame) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERegion &region, const eVision::EFrame &frame, uint32_t topologyLineIndex, bool automaticReadingDirection) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
eVision::EasyDeepOCR::ETextLine	Recognize (const eVision::EBaseROI &roi, const eVision::ERegion &region, const eVision::EFrame &frame, uint32_t topologyLineIndex) const
	Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).
std::vector< eVision::EasyDeepOCR::EText >	Merge (const std::vector< eVision::EasyDeepOCR::ETextLine > &textlines) const
	Merges geometrically consistent ETextLines into structured ETexts satisfying the specified ETextReader::SetTopology. Throws if no ETextReader::SetTopology has been specified.
	ETextReader (CppWrapperInternal::InternalConstructor)

Public Attributes
CppWrapperInternal::WrappedClass	wrappedClass_

Detailed Description

A ETextReader aims to find and read texts inside a given ROI, through ETextReader::Read. The returned texts can be filtered by providing a ETopology. An optimization phase can be performed before starting to read, to increase the processing speed of ETextReader. This optimization phase can be launched with ETextReader::Optimize.

License(s): EasyDeepOCR

Constructor & Destructor Documentation

◆ ETextReader() [1/3]

eVision::EasyDeepOCR::ETextReader::ETextReader ( )

Creates a ETextReader.

◆ ETextReader() [2/3]

eVision::EasyDeepOCR::ETextReader::ETextReader ( const eVision::EasyDeepOCR::ETextReader & other )

Creates a ETextReader.

Parameters

other Another ETextReader to be copied in the new ETextReader.

◆ ~ETextReader()

eVision::EasyDeepOCR::ETextReader::~ETextReader	(		)
throw	(	)

virtual

◆ ETextReader() [3/3]

eVision::EasyDeepOCR::ETextReader::ETextReader ( CppWrapperInternal::InternalConstructor )

Member Function Documentation

◆ GetExecutionSettings()

eVision::EasyDeepLearning::EDeepLearningExecutionSettings eVision::EasyDeepOCR::ETextReader::GetExecutionSettings ( ) const

EDeepLearningExecutionSettings used by the ETextReader.

◆ GetHasLocationCustomModel()

bool eVision::EasyDeepOCR::ETextReader::GetHasLocationCustomModel ( ) const

Whether the reader has a location custom model.

◆ GetHasOrientationCustomModel()

bool eVision::EasyDeepOCR::ETextReader::GetHasOrientationCustomModel ( ) const

Whether the reader has a orientation custom model.

◆ GetHasRecognitionCustomModel()

bool eVision::EasyDeepOCR::ETextReader::GetHasRecognitionCustomModel ( ) const

Whether the reader has a recognition custom model.

◆ GetHasTextAngle()

bool eVision::EasyDeepOCR::ETextReader::GetHasTextAngle ( ) const

Returns true if the text angle hint is defined.

◆ GetHasTopology()

bool eVision::EasyDeepOCR::ETextReader::GetHasTopology ( ) const

Returns true if the ETopology is set.

◆ GetTextAngle()

float eVision::EasyDeepOCR::ETextReader::GetTextAngle ( ) const

Angle hint for the text to be found.

◆ GetTextLineHeight()

eVision::EFloatRange eVision::EasyDeepOCR::ETextReader::GetTextLineHeight ( ) const

Gets the range for the text line height. Throws if not previously set. By default, it is not set, so no filtering is applied.

◆ GetTopology()

eVision::ETopology eVision::EasyDeepOCR::ETextReader::GetTopology ( ) const

ETopology of the ETextReader.

◆ HasLocationCustomModel()

bool eVision::EasyDeepOCR::ETextReader::HasLocationCustomModel ( ) const

Whether the reader has a location custom model.

Deprecated

Use ETextReader::GetHasLocationCustomModel instead.

Deprecated in version 26.10. Scheduled to be removed in version 27.10.

◆ HasOrientationCustomModel()

bool eVision::EasyDeepOCR::ETextReader::HasOrientationCustomModel ( ) const

Whether the reader has a orientation custom model.

Deprecated

Use ETextReader::GetHasOrientationCustomModel instead.

Deprecated in version 26.10. Scheduled to be removed in version 27.10.

◆ HasRecognitionCustomModel()

bool eVision::EasyDeepOCR::ETextReader::HasRecognitionCustomModel ( ) const

Whether the reader has a recognition custom model.

Deprecated

Use ETextReader::GetHasRecognitionCustomModel instead.

Deprecated in version 26.10. Scheduled to be removed in version 27.10.

◆ Initialize()

void eVision::EasyDeepOCR::ETextReader::Initialize ( )

Initializes the ETextReader. ETextReader must be initialized before using ETextReader::Read.

◆ IsEqual()

bool eVision::EasyDeepOCR::ETextReader::IsEqual ( const eVision::EasyDeepOCR::ETextReader & other ) const

Checks if this ETextReader instance is strictly equal to another ETextReader instance.

Parameters

other Reference to the ETextReader instance to compare with

◆ IsInitialized()

bool eVision::EasyDeepOCR::ETextReader::IsInitialized ( ) const

Returns true if the ETextReader is initialized.

◆ IsOptimized()

bool eVision::EasyDeepOCR::ETextReader::IsOptimized ( ) const

Returns true if the ETextReader has been optimized with a call to ETextReader::Optimize, and false otherwise.

◆ Load() [1/2]

void eVision::EasyDeepOCR::ETextReader::Load ( const std::string & path )

Loads the ETextReader.

Parameters

path	Path where to save the ETextReader.

◆ Load() [2/2]

void eVision::EasyDeepOCR::ETextReader::Load ( eVision::ESerializer * serializer )

Loads the ETextReader.

Parameters

serializer

◆ Localize() [1/4]

std::vector< eVision::ERectangle > eVision::EasyDeepOCR::ETextReader::Localize ( const eVision::EBaseROI & roi ) const

Localizes the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.

Parameters

roi	The region of interest (ROI) in which to search for text. Only EROIBW8 and EROIC24 are supported.

Deprecated

Use ETextReader::Locate instead.

Deprecated in version 26.6. Scheduled to be removed in version 27.6.

◆ Localize() [2/4]

std::vector< eVision::ERectangle > eVision::EasyDeepOCR::ETextReader::Localize	(	const eVision::EBaseROI &	roi,
		const eVision::EFrame &	frame ) const

Localizes the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.

Parameters

roi	The region of interest (ROI) in which to search for text. Only EROIBW8 and EROIC24 are supported.
frame	Reference frame of the input region.

Deprecated

Use ETextReader::Locate instead.

Deprecated in version 26.6. Scheduled to be removed in version 27.6.

◆ Localize() [3/4]

std::vector< eVision::ERectangle > eVision::EasyDeepOCR::ETextReader::Localize	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region ) const

Localizes the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.

Parameters

roi	The region of interest (ROI) in which to search for text. Only EROIBW8 and EROIC24 are supported.
region	Optional parameter that specifies a sub-region within the ROI to limit the text search area.

Deprecated

Use ETextReader::Locate instead.

Deprecated in version 26.6. Scheduled to be removed in version 27.6.

◆ Localize() [4/4]

std::vector< eVision::ERectangle > eVision::EasyDeepOCR::ETextReader::Localize	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region,
		const eVision::EFrame &	frame ) const

Localizes the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.

Parameters

roi	The region of interest (ROI) in which to search for text. Only EROIBW8 and EROIC24 are supported.
region	Optional parameter that specifies a sub-region within the ROI to limit the text search area.
frame	Reference frame of the input region.

Deprecated

Use ETextReader::Locate instead.

Deprecated in version 26.6. Scheduled to be removed in version 27.6.

◆ Locate() [1/4]

std::vector< eVision::ERectangle > eVision::EasyDeepOCR::ETextReader::Locate ( const eVision::EBaseROI & roi ) const

Locates the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.

Parameters

roi	The region of interest (ROI) in which to search for text. Only EROIBW8 and EROIC24 are supported.

◆ Locate() [2/4]

std::vector< eVision::ERectangle > eVision::EasyDeepOCR::ETextReader::Locate	(	const eVision::EBaseROI &	roi,
		const eVision::EFrame &	frame ) const

Locates the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.

Parameters

roi	The region of interest (ROI) in which to search for text. Only EROIBW8 and EROIC24 are supported.
frame	Reference frame of the input region.

◆ Locate() [3/4]

std::vector< eVision::ERectangle > eVision::EasyDeepOCR::ETextReader::Locate	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region ) const

Locates the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.

Parameters

roi	The region of interest (ROI) in which to search for text. Only EROIBW8 and EROIC24 are supported.
region	Optional parameter that specifies a sub-region within the ROI to limit the text search area.

◆ Locate() [4/4]

std::vector< eVision::ERectangle > eVision::EasyDeepOCR::ETextReader::Locate	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region,
		const eVision::EFrame &	frame ) const

Locates the texts within a specified region of interest (ROI). Returns the detected text areas as ERectangles.

Parameters

roi	The region of interest (ROI) in which to search for text. Only EROIBW8 and EROIC24 are supported.
region	Optional parameter that specifies a sub-region within the ROI to limit the text search area.
frame	Reference frame of the input region.

◆ Merge()

std::vector< eVision::EasyDeepOCR::EText > eVision::EasyDeepOCR::ETextReader::Merge ( const std::vector< eVision::EasyDeepOCR::ETextLine > & textlines ) const

Merges geometrically consistent ETextLines into structured ETexts satisfying the specified ETextReader::SetTopology. Throws if no ETextReader::SetTopology has been specified.

Parameters

textlines A collection of ETextLine candidates to be merged.

◆ operator=()

eVision::EasyDeepOCR::ETextReader & eVision::EasyDeepOCR::ETextReader::operator= ( const eVision::EasyDeepOCR::ETextReader & other )

Assignment operator.

Parameters

other Another ETextReader to be assigned in the new ETextReader.

◆ operator==()

bool eVision::EasyDeepOCR::ETextReader::operator== ( const eVision::EasyDeepOCR::ETextReader & other ) const

Checks if this ETextReader instance is strictly equal to another ETextReader instance.

Parameters

other Reference to the ETextReader instance to compare with

This operator is stricly equivalent to the ETextReader::IsEqual method.

◆ Optimize() [1/3]

eVision::EasyDeepOCR::EOptimizationReport eVision::EasyDeepOCR::ETextReader::Optimize ( const eVision::EasyDeepOCR::EOptimizationReport & optimizationReport )

Sets the optimal internal parameters by finding the text in the corresponding input image, for each ESample.

Parameters

optimizationReport Optimization report to be restored.

Once an optimization is performed, the ETextReader is specialized and only returns texts sharing geometrical properties with the ones it has learned from.

◆ Optimize() [2/3]

eVision::EasyDeepOCR::EOptimizationReport eVision::EasyDeepOCR::ETextReader::Optimize ( const eVision::EasyDeepOCR::ESampleList & samples )

Sets the optimal internal parameters by finding the text in the corresponding input image, for each ESample.

Parameters

samples Samples to be optimized from.

Once an optimization is performed, the ETextReader is specialized and only returns texts sharing geometrical properties with the ones it has learned from.

◆ Optimize() [3/3]

eVision::EasyDeepOCR::EOptimizationReport eVision::EasyDeepOCR::ETextReader::Optimize	(	const eVision::EasyDeepOCR::ESampleList &	samples,
		eVision::EImageType	imageType )

Sets the optimal internal parameters by finding the text in the corresponding input image, for each ESample.

Parameters

samples	Samples to be optimized from.
imageType	Specifies the image type for loading the sample images. If not provided, the image type will be automatically determined based on the sample image files.

Once an optimization is performed, the ETextReader is specialized and only returns texts sharing geometrical properties with the ones it has learned from.

◆ Read() [1/4]

std::vector< eVision::EasyDeepOCR::EText > eVision::EasyDeepOCR::ETextReader::Read ( const eVision::EBaseROI & roi ) const

Reads texts inside an EBaseROI. Texts are returned as EText. Texts comply with the ETopology previously specified with ETextReader::SetTopology, if any. The processing speed can be increased with ETextReader::Optimize.

Parameters

roi	ROI in which the texts have to be found. Only EROIBW8 and EROIC24 are supported.

◆ Read() [2/4]

std::vector< eVision::EasyDeepOCR::EText > eVision::EasyDeepOCR::ETextReader::Read	(	const eVision::EBaseROI &	roi,
		const eVision::EFrame &	frame ) const

Reads texts inside an EBaseROI. Texts are returned as EText. Texts comply with the ETopology previously specified with ETextReader::SetTopology, if any. The processing speed can be increased with ETextReader::Optimize.

Parameters

roi	ROI in which the texts have to be found. Only EROIBW8 and EROIC24 are supported.
frame	Reference frame of the input region.

◆ Read() [3/4]

std::vector< eVision::EasyDeepOCR::EText > eVision::EasyDeepOCR::ETextReader::Read	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region ) const

Reads texts inside an EBaseROI. Texts are returned as EText. Texts comply with the ETopology previously specified with ETextReader::SetTopology, if any. The processing speed can be increased with ETextReader::Optimize.

Parameters

roi	ROI in which the texts have to be found. Only EROIBW8 and EROIC24 are supported.
region	Region in which the texts have to be found.

◆ Read() [4/4]

std::vector< eVision::EasyDeepOCR::EText > eVision::EasyDeepOCR::ETextReader::Read	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region,
		const eVision::EFrame &	frame ) const

Reads texts inside an EBaseROI. Texts are returned as EText. Texts comply with the ETopology previously specified with ETextReader::SetTopology, if any. The processing speed can be increased with ETextReader::Optimize.

Parameters

roi	ROI in which the texts have to be found. Only EROIBW8 and EROIC24 are supported.
region	Region in which the texts have to be found.
frame	Reference frame of the input region.

◆ Recognize() [1/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize ( const eVision::EBaseROI & miniature ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

miniature The ROI containing the miniature image with text to recognize.

◆ Recognize() [2/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	miniature,
		bool	automaticReadingDirection ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

miniature	The ROI containing the miniature image with text to recognize.
automaticReadingDirection	When true, automatically detects the reading direction of the text. Default is false.

◆ Recognize() [3/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	miniature,
		uint32_t	topologyLineIndex ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

miniature	The ROI containing the miniature image with text to recognize.
topologyLineIndex	The index of the ETopologyLine to use for text recognition constraints.

◆ Recognize() [4/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	miniature,
		uint32_t	topologyLineIndex,
		bool	automaticReadingDirection ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

miniature	The ROI containing the miniature image with text to recognize.
topologyLineIndex	The index of the ETopologyLine to use for text recognition constraints.
automaticReadingDirection	When true, automatically detects the reading direction of the text. Default is false.

◆ Recognize() [5/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERectangle &	rectangle ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
rectangle	Defines the area within the source ROI to extract as the miniature image.

◆ Recognize() [6/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERectangle &	rectangle,
		bool	automaticReadingDirection ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
rectangle	Defines the area within the source ROI to extract as the miniature image.
automaticReadingDirection	When true, automatically detects the reading direction of the text. Default is false.

◆ Recognize() [7/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERectangle &	rectangle,
		const eVision::EFrame &	frame ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
rectangle	Defines the area within the source ROI to extract as the miniature image.
frame	Reference frame of the input region.

◆ Recognize() [8/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERectangle &	rectangle,
		const eVision::EFrame &	frame,
		bool	automaticReadingDirection ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
rectangle	Defines the area within the source ROI to extract as the miniature image.
frame	Reference frame of the input region.
automaticReadingDirection	When true, automatically detects the reading direction of the text. Default is false.

◆ Recognize() [9/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERectangle &	rectangle,
		const eVision::EFrame &	frame,
		uint32_t	topologyLineIndex ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
rectangle	Defines the area within the source ROI to extract as the miniature image.
frame	Reference frame of the input region.
topologyLineIndex	The index of the ETopologyLine to use for text recognition constraints.

◆ Recognize() [10/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERectangle &	rectangle,
		const eVision::EFrame &	frame,
		uint32_t	topologyLineIndex,
		bool	automaticReadingDirection ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
rectangle	Defines the area within the source ROI to extract as the miniature image.
frame	Reference frame of the input region.
topologyLineIndex	The index of the ETopologyLine to use for text recognition constraints.
automaticReadingDirection	When true, automatically detects the reading direction of the text. Default is false.

◆ Recognize() [11/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERectangle &	rectangle,
		uint32_t	topologyLineIndex ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
rectangle	Defines the area within the source ROI to extract as the miniature image.
topologyLineIndex	The index of the ETopologyLine to use for text recognition constraints.

◆ Recognize() [12/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERectangle &	rectangle,
		uint32_t	topologyLineIndex,
		bool	automaticReadingDirection ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
rectangle	Defines the area within the source ROI to extract as the miniature image.
topologyLineIndex	The index of the ETopologyLine to use for text recognition constraints.
automaticReadingDirection	When true, automatically detects the reading direction of the text. Default is false.

◆ Recognize() [13/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
region	Defines the area within the source ROI to extract as the miniature image.

◆ Recognize() [14/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region,
		bool	automaticReadingDirection ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
region	Defines the area within the source ROI to extract as the miniature image.
automaticReadingDirection	When true, automatically detects the reading direction of the text. Default is false.

◆ Recognize() [15/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region,
		const eVision::EFrame &	frame ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
region	Defines the area within the source ROI to extract as the miniature image.
frame	Reference frame of the input region.

◆ Recognize() [16/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region,
		const eVision::EFrame &	frame,
		bool	automaticReadingDirection ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
region	Defines the area within the source ROI to extract as the miniature image.
frame	Reference frame of the input region.
automaticReadingDirection	When true, automatically detects the reading direction of the text. Default is false.

◆ Recognize() [17/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region,
		const eVision::EFrame &	frame,
		uint32_t	topologyLineIndex ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
region	Defines the area within the source ROI to extract as the miniature image.
frame	Reference frame of the input region.
topologyLineIndex	The index of the ETopologyLine to use for text recognition constraints.

◆ Recognize() [18/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region,
		const eVision::EFrame &	frame,
		uint32_t	topologyLineIndex,
		bool	automaticReadingDirection ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
region	Defines the area within the source ROI to extract as the miniature image.
frame	Reference frame of the input region.
topologyLineIndex	The index of the ETopologyLine to use for text recognition constraints.
automaticReadingDirection	When true, automatically detects the reading direction of the text. Default is false.

◆ Recognize() [19/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region,
		uint32_t	topologyLineIndex ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
region	Defines the area within the source ROI to extract as the miniature image.
topologyLineIndex	The index of the ETopologyLine to use for text recognition constraints.

◆ Recognize() [20/20]

eVision::EasyDeepOCR::ETextLine eVision::EasyDeepOCR::ETextReader::Recognize	(	const eVision::EBaseROI &	roi,
		const eVision::ERegion &	region,
		uint32_t	topologyLineIndex,
		bool	automaticReadingDirection ) const

Recognizes a text within a miniature image and returns it as an ETextLine. The returned text will conform to the ETopologyLine if one is provided via the topologyLineIndex parameter. The miniature image can be provided directly or extracted from a larger ROI using an ERectangle or an ERegion. If the text in the miniature image is not oriented along the standard reading direction, setting automaticReadingDirection to true will attempt to detect and adjust for the correct reading orientation. When ETextReader::GetHasTextAngle is true, automaticReadingDirection is ignored and the reading orientation is chosen as the closest to ETextReader::SetTextAngle. Otherewise, To ensure the oriented rectangle obtained from the region is deterministic, the reader will always snap the orientation to the horizontal axis (changing the width and height accordingly).

Parameters

roi	The source ROI from which to extract the miniature image.
region	Defines the area within the source ROI to extract as the miniature image.
topologyLineIndex	The index of the ETopologyLine to use for text recognition constraints.
automaticReadingDirection	When true, automatically detects the reading direction of the text. Default is false.

◆ ResetLocationCustomModel()

void eVision::EasyDeepOCR::ETextReader::ResetLocationCustomModel ( )

Removes any location custom model that was previously set and use the default location model.

◆ ResetOptimization()

void eVision::EasyDeepOCR::ETextReader::ResetOptimization ( )

Forgets the optimal parameters previously set by ETextReader::Optimize.

◆ ResetOrientationCustomModel()

void eVision::EasyDeepOCR::ETextReader::ResetOrientationCustomModel ( )

Removes any orientation custom model that was previously set and use the default orientation model.

◆ ResetRecognitionCustomModel()

void eVision::EasyDeepOCR::ETextReader::ResetRecognitionCustomModel ( )

Removes any recognition custom model that was previously set and use the default recognition model.

◆ ResetTopology()

void eVision::EasyDeepOCR::ETextReader::ResetTopology ( )

Forgets the ETopology previously set.

◆ Save() [1/2]

void eVision::EasyDeepOCR::ETextReader::Save ( const std::string & path )

Saves the ETextReader.

Parameters

path	Path where to save the ETextReader.

◆ Save() [2/2]

void eVision::EasyDeepOCR::ETextReader::Save ( eVision::ESerializer * serializer )

Saves the ETextReader.

Parameters

serializer

◆ SetExecutionSettings()

void eVision::EasyDeepOCR::ETextReader::SetExecutionSettings ( const eVision::EasyDeepLearning::EDeepLearningExecutionSettings & executionSettings )

EDeepLearningExecutionSettings used by the ETextReader.

◆ SetLocationCustomModel()

void eVision::EasyDeepOCR::ETextReader::SetLocationCustomModel ( const eVision::EasyDeepLearning::EDeepOCRTool & tool )

Sets a location custom model from a EDeepOCRTool object (user trained model).

Parameters

tool	Trained EDeepOCRTool tool with EDeepOCRModel::EDeepOCRModel_Location model

Note that the location custom model will be serialized as part of the text reader.

◆ SetLocationCustomModelPath()

void eVision::EasyDeepOCR::ETextReader::SetLocationCustomModelPath ( const std::string & modelPath )

Sets a location custom model using the path to a saved EDeepOCRTool (user trained model).

Parameters

modelPath Path to a EDeepOCRTool tool.

Note that the location custom model will be serialized as part of the text reader.

◆ SetOrientationCustomModel()

void eVision::EasyDeepOCR::ETextReader::SetOrientationCustomModel ( const eVision::EasyDeepLearning::EDeepOCRTool & tool )

Sets an orientation custom model from a EDeepOCRTool object (user trained model).

Parameters

tool	Trained EDeepOCRTool tool with EDeepOCRModel::EDeepOCRModel_Orientation model

Note that the orientation custom model will be serialized as part of the text reader.

◆ SetOrientationCustomModelPath()

void eVision::EasyDeepOCR::ETextReader::SetOrientationCustomModelPath ( const std::string & modelPath )

Sets an orientation custom model using the path to a saved EDeepOCRTool (user trained model).

Parameters

modelPath Path to a EDeepOCRTool tool.

Note that the orientation custom model will be serialized as part of the text reader.

◆ SetRecognitionCustomModel()

void eVision::EasyDeepOCR::ETextReader::SetRecognitionCustomModel ( const eVision::EasyDeepLearning::EDeepOCRTool & tool )

Sets a recognition custom model from a EDeepOCRTool object (user trained model).

Parameters

tool	Trained EDeepOCRTool tool with EDeepOCRModel::EDeepOCRModel_Recognition model

Note that the recognition custom model will be serialized as part of the text reader.

◆ SetRecognitionCustomModelPath()

void eVision::EasyDeepOCR::ETextReader::SetRecognitionCustomModelPath ( const std::string & modelPath )

Sets a recognition custom model using the path to a saved EDeepOCRTool (user trained model).

Parameters

modelPath Path to a EDeepOCRTool tool.

Note that the recognition custom model will be serialized as part of the text reader.

◆ SetTextAngle()

void eVision::EasyDeepOCR::ETextReader::SetTextAngle ( float arg0 )

Angle hint for the text to be found.

◆ SetTextLineHeight()

void eVision::EasyDeepOCR::ETextReader::SetTextLineHeight	(	int	minHeight,
		int	maxHeight )

Sets the range for the text line height. All returned text lines will have a height in the range [minHeight, maxHeight]. Any text line that does not fit this constraint, will be discarded. By default, it is not set, so no filtering is applied.

Parameters

minHeight	Lower bound for the line height range.
maxHeight	Upper bound for the line height range.

◆ SetTopology()

void eVision::EasyDeepOCR::ETextReader::SetTopology ( const eVision::ETopology & topology )

ETopology of the ETextReader.

◆ UnsetTextAngle()

void eVision::EasyDeepOCR::ETextReader::UnsetTextAngle ( )

Unsets the text angle hint.

Member Data Documentation

◆ wrappedClass_

CppWrapperInternal::WrappedClass eVision::EasyDeepOCR::ETextReader::wrappedClass_

Public Member Functions

Public Attributes

Detailed Description

Constructor & Destructor Documentation

◆ ETextReader() [1/3]

◆ ETextReader() [2/3]

◆ ~ETextReader()

◆ ETextReader() [3/3]

Member Function Documentation

◆ GetExecutionSettings()

◆ GetHasLocationCustomModel()

◆ GetHasOrientationCustomModel()

◆ GetHasRecognitionCustomModel()

◆ GetHasTextAngle()

◆ GetHasTopology()

◆ GetTextAngle()

◆ GetTextLineHeight()

◆ GetTopology()

◆ HasLocationCustomModel()

◆ HasOrientationCustomModel()

◆ HasRecognitionCustomModel()

◆ Initialize()

◆ IsEqual()

◆ IsInitialized()

◆ IsOptimized()

◆ Load() [1/2]

◆ Load() [2/2]

◆ Localize() [1/4]

◆ Localize() [2/4]

◆ Localize() [3/4]

◆ Localize() [4/4]

◆ Locate() [1/4]

◆ Locate() [2/4]

◆ Locate() [3/4]

◆ Locate() [4/4]

◆ Merge()

◆ operator=()

◆ operator==()

◆ Optimize() [1/3]

◆ Optimize() [2/3]

◆ Optimize() [3/3]

◆ Read() [1/4]

◆ Read() [2/4]

◆ Read() [3/4]

◆ Read() [4/4]

◆ Recognize() [1/20]

◆ Recognize() [2/20]

◆ Recognize() [3/20]

◆ Recognize() [4/20]

◆ Recognize() [5/20]

◆ Recognize() [6/20]

◆ Recognize() [7/20]

◆ Recognize() [8/20]

◆ Recognize() [9/20]

◆ Recognize() [10/20]

◆ Recognize() [11/20]

◆ Recognize() [12/20]

◆ Recognize() [13/20]

◆ Recognize() [14/20]

◆ Recognize() [15/20]

◆ Recognize() [16/20]

◆ Recognize() [17/20]

◆ Recognize() [18/20]

◆ Recognize() [19/20]

◆ Recognize() [20/20]

◆ ResetLocationCustomModel()

◆ ResetOptimization()

◆ ResetOrientationCustomModel()

◆ ResetRecognitionCustomModel()

◆ ResetTopology()

◆ Save() [1/2]

◆ Save() [2/2]

◆ SetExecutionSettings()

◆ SetLocationCustomModel()

◆ SetLocationCustomModelPath()

◆ SetOrientationCustomModel()

◆ SetOrientationCustomModelPath()

◆ SetRecognitionCustomModel()

◆ SetRecognitionCustomModelPath()

◆ SetTextAngle()