Manages a complete context for the EChecker golden template inspection tool. More...

#include <eVision/EasyObject/EChecker/EChecker.h>

Collaboration diagram for eVision::EChecker:

Public Member Functions
	EChecker ()
	Constructs a golden template inspection context.
	EChecker (const eVision::EChecker &other)
	Constructs a golden template inspection context.
virtual	~EChecker () throw ()
eVision::EChecker &	operator= (const eVision::EChecker &other)
	Assignment operator.
void	SetTrainingMode (eVision::ETrainingMode mode)
	Training mode. The training mode determines which process will be used to build the threshold images used for inspection. EChecker has two training modes: A Quick mode, efficient for crude defects and stable images, and a Precise mode, designed for more subtle defects and handling more variability in the images. Default: Precise training mode.
eVision::ETrainingMode	GetTrainingMode () const
	Training mode. The training mode determines which process will be used to build the threshold images used for inspection. EChecker has two training modes: A Quick mode, efficient for crude defects and stable images, and a Precise mode, designed for more subtle defects and handling more variability in the images. Default: Precise training mode.
eVision::ENormalizationMode	GetNormalizationMode () const
	Normalization mode. This setting allows you to specify if you want to enable normalization, and if yes, which type. Normalization allows the training and inspection processes to automatically correct global illumination differences between the images. Default: Moments based normalization.
void	SetNormalizationMode (eVision::ENormalizationMode mode)
	Normalization mode. This setting allows you to specify if you want to enable normalization, and if yes, which type. Normalization allows the training and inspection processes to automatically correct global illumination differences between the images. Default: Moments based normalization.
void	Train (const std::vector< const eVision::EROIBW8 * > &rois, const std::vector< eVision::EFrame > &frames, const eVision::ERegion &inspectionRegion)
	Train. The distribution of noise is estimated for each pixel based on the training images. At inspection, a pixel is validated it if is close enough to the average value on the training set. The size of that interval is proportional to EChecker::GetInspectionTolerance.
void	Train (const std::vector< const eVision::EROIBW8 * > &rois, const std::vector< eVision::EFrame > &frames)
	Train. The distribution of noise is estimated for each pixel based on the training images. At inspection, a pixel is validated it if is close enough to the average value on the training set. The size of that interval is proportional to EChecker::GetInspectionTolerance.
void	Train (const std::vector< const eVision::EROIBW8 * > &rois, const eVision::ERegion &inspectionRegion)
	Train. The distribution of noise is estimated for each pixel based on the training images. At inspection, a pixel is validated it if is close enough to the average value on the training set. The size of that interval is proportional to EChecker::GetInspectionTolerance.
void	Train (const std::vector< const eVision::EROIBW8 * > &rois)
	Train. The distribution of noise is estimated for each pixel based on the training images. At inspection, a pixel is validated it if is close enough to the average value on the training set. The size of that interval is proportional to EChecker::GetInspectionTolerance.
void	Train (const std::vector< eVision::EROIBW8 > &rois, const std::vector< eVision::EFrame > &frames, const eVision::ERegion &inspectionRegion)
	Train. The distribution of noise is estimated for each pixel based on the training images. At inspection, a pixel is validated it if is close enough to the average value on the training set. The size of that interval is proportional to EChecker::GetInspectionTolerance.
void	Train (const std::vector< eVision::EROIBW8 > &rois, const std::vector< eVision::EFrame > &frames)
	Train. The distribution of noise is estimated for each pixel based on the training images. At inspection, a pixel is validated it if is close enough to the average value on the training set. The size of that interval is proportional to EChecker::GetInspectionTolerance.
void	Train (const std::vector< eVision::EROIBW8 > &rois, const eVision::ERegion &inspectionRegion)
	Train. The distribution of noise is estimated for each pixel based on the training images. At inspection, a pixel is validated it if is close enough to the average value on the training set. The size of that interval is proportional to EChecker::GetInspectionTolerance.
void	Train (const std::vector< eVision::EROIBW8 > &rois)
	Train. The distribution of noise is estimated for each pixel based on the training images. At inspection, a pixel is validated it if is close enough to the average value on the training set. The size of that interval is proportional to EChecker::GetInspectionTolerance.
void	Train (const std::vector< eVision::EImageBW8 > &rois, const std::vector< eVision::EFrame > &frames, const eVision::ERegion &inspectionRegion)
	Train. The distribution of noise is estimated for each pixel based on the training images. At inspection, a pixel is validated it if is close enough to the average value on the training set. The size of that interval is proportional to EChecker::GetInspectionTolerance.
void	Train (const std::vector< eVision::EImageBW8 > &rois, const std::vector< eVision::EFrame > &frames)
	Train. The distribution of noise is estimated for each pixel based on the training images. At inspection, a pixel is validated it if is close enough to the average value on the training set. The size of that interval is proportional to EChecker::GetInspectionTolerance.
void	Train (const std::vector< eVision::EImageBW8 > &rois, const eVision::ERegion &inspectionRegion)
	Train. The distribution of noise is estimated for each pixel based on the training images. At inspection, a pixel is validated it if is close enough to the average value on the training set. The size of that interval is proportional to EChecker::GetInspectionTolerance.
void	Train (const std::vector< eVision::EImageBW8 > &rois)
	Train. The distribution of noise is estimated for each pixel based on the training images. At inspection, a pixel is validated it if is close enough to the average value on the training set. The size of that interval is proportional to EChecker::GetInspectionTolerance.
void	TrainOnSingleImage (const eVision::EROIBW8 &roi, eVision::EFrame &frame, const eVision::ERegion &inspectionRegion, uint32_t geometricTolerance, uint32_t grayLevelTolerance)
	Train on a single image. At inspection, a pixel is validated if it is above "the smallest pixel in a geometricTolerance vicinity" - grayLevelTolerance and below "the biggest pixel in a geometricTolerance vicinity" + grayLevelTolerance. EChecker::GetInspectionTolerance and EChecker::SetTrainingMode are ignored by this method.
void	TrainOnSingleImage (const eVision::EROIBW8 &roi, eVision::EFrame &frame, uint32_t geometricTolerance, uint32_t grayLevelTolerance)
	Train on a single image. At inspection, a pixel is validated if it is above "the smallest pixel in a geometricTolerance vicinity" - grayLevelTolerance and below "the biggest pixel in a geometricTolerance vicinity" + grayLevelTolerance. EChecker::GetInspectionTolerance and EChecker::SetTrainingMode are ignored by this method.
void	TrainOnSingleImage (const eVision::EROIBW8 &roi, const eVision::ERegion &inspectionRegion, uint32_t geometricTolerance, uint32_t grayLevelTolerance)
	Train on a single image. At inspection, a pixel is validated if it is above "the smallest pixel in a geometricTolerance vicinity" - grayLevelTolerance and below "the biggest pixel in a geometricTolerance vicinity" + grayLevelTolerance. EChecker::GetInspectionTolerance and EChecker::SetTrainingMode are ignored by this method.
void	TrainOnSingleImage (const eVision::EROIBW8 &roi, uint32_t geometricTolerance, uint32_t grayLevelTolerance)
	Train on a single image. At inspection, a pixel is validated if it is above "the smallest pixel in a geometricTolerance vicinity" - grayLevelTolerance and below "the biggest pixel in a geometricTolerance vicinity" + grayLevelTolerance. EChecker::GetInspectionTolerance and EChecker::SetTrainingMode are ignored by this method.
void	ResetTraining ()
	Reset training. Use this if you want to reset the state of the object.
float	GetInspectionTolerance () const
	Inspection Tolerance factor. If you need to be more tolerant towards noise, texture or illumination irregularities, raise this value. If you have clean images, low texture and some defects are missed, lower this value. Default value: 4.0.
void	SetInspectionTolerance (float tolerance)
	Inspection Tolerance factor. If you need to be more tolerant towards noise, texture or illumination irregularities, raise this value. If you have clean images, low texture and some defects are missed, lower this value. Default value: 4.0.
void	Inspect (const eVision::EROIBW8 &roi, const eVision::EFrame &frame, eVision::ECodedImage2 &defects)
	Inspect the image. The defects are returned in the referential images aligned on the inspection region. See EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage and EChecker::GetLastRegisteredImage.
void	Inspect (const eVision::EROIBW8 &roi, eVision::ECodedImage2 &defects)
	Inspect the image. The defects are returned in the referential images aligned on the inspection region. See EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage and EChecker::GetLastRegisteredImage.
const eVision::EImageBW8 &	GetLowThresholdImage () const
	Low threshold image.
const eVision::EImageBW8 &	GetHighThresholdImage () const
	High threshold image.
const eVision::EImageBW8 &	GetLastRegisteredImage () const
	Represents the last source image, after it has been aligned with the reference image and normalized.
const eVision::ERegion &	GetInspectionRegion () const
	Represents the inspection region, overlayed on EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage or EChecker::GetLastRegisteredImage.
bool	GetIsTrained () const
	Application state. Has the checker been trained.
void	DrawInspectionRegion (eVision::EDrawAdapter *graphicContext, float zoomX, float zoomY, float panX, float panY) const
	Draws the inspection region on EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage or EChecker::GetLastRegisteredImage.
void	DrawInspectionRegion (eVision::EDrawAdapter *graphicContext, float zoomX, float zoomY, float panX) const
	Draws the inspection region on EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage or EChecker::GetLastRegisteredImage.
void	DrawInspectionRegion (eVision::EDrawAdapter *graphicContext, float zoomX, float zoomY) const
	Draws the inspection region on EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage or EChecker::GetLastRegisteredImage.
void	DrawInspectionRegion (eVision::EDrawAdapter *graphicContext, float zoomX) const
	Draws the inspection region on EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage or EChecker::GetLastRegisteredImage.
void	DrawInspectionRegion (eVision::EDrawAdapter *graphicContext) const
	Draws the inspection region on EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage or EChecker::GetLastRegisteredImage.
void	Load (const std::string &file)
	Load an inspection model.
void	Load (eVision::ESerializer *serializer)
	Load an inspection model.
void	Save (eVision::ESerializer *serializer)
	Save the inspection model.
void	Save (const std::string &file)
	Save the inspection model.
	EChecker (CppWrapperInternal::InternalConstructor)

Public Attributes
CppWrapperInternal::WrappedClass	wrappedClass_

Detailed Description

Manages a complete context for the EChecker golden template inspection tool.

License(s): EasyObject

Constructor & Destructor Documentation

◆ EChecker() [1/3]

eVision::EChecker::EChecker ( )

Constructs a golden template inspection context.

◆ EChecker() [2/3]

eVision::EChecker::EChecker ( const eVision::EChecker & other )

Constructs a golden template inspection context.

Parameters

other The EChecker to copy into this one.

◆ ~EChecker()

eVision::EChecker::~EChecker	(		)
throw	(	)

virtual

◆ EChecker() [3/3]

eVision::EChecker::EChecker ( CppWrapperInternal::InternalConstructor )

Member Function Documentation

◆ DrawInspectionRegion() [1/5]

void eVision::EChecker::DrawInspectionRegion ( eVision::EDrawAdapter * graphicContext ) const

Draws the inspection region on EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage or EChecker::GetLastRegisteredImage.

Parameters

graphicContext Device context of the drawing window.

◆ DrawInspectionRegion() [2/5]

void eVision::EChecker::DrawInspectionRegion	(	eVision::EDrawAdapter *	graphicContext,
		float	zoomX ) const

Draws the inspection region on EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage or EChecker::GetLastRegisteredImage.

Parameters

graphicContext	Device context of the drawing window.
zoomX	Magnification factor for zooming in or out in the horizontal direction. By default, the image is displayed in true scale.

◆ DrawInspectionRegion() [3/5]

void eVision::EChecker::DrawInspectionRegion	(	eVision::EDrawAdapter *	graphicContext,
		float	zoomX,
		float	zoomY ) const

Draws the inspection region on EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage or EChecker::GetLastRegisteredImage.

Parameters

graphicContext	Device context of the drawing window.
zoomX	Magnification factor for zooming in or out in the horizontal direction. By default, the image is displayed in true scale.
zoomY	Magnification factor for zooming in or out in the vertical direction. Setting a 0 value (which is the default) will result in isotropic scaling (i.e. equal horizontal and vertical factors).

◆ DrawInspectionRegion() [4/5]

void eVision::EChecker::DrawInspectionRegion	(	eVision::EDrawAdapter *	graphicContext,
		float	zoomX,
		float	zoomY,
		float	panX ) const

Draws the inspection region on EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage or EChecker::GetLastRegisteredImage.

Parameters

graphicContext	Device context of the drawing window.
zoomX	Magnification factor for zooming in or out in the horizontal direction. By default, the image is displayed in true scale.
zoomY	Magnification factor for zooming in or out in the vertical direction. Setting a 0 value (which is the default) will result in isotropic scaling (i.e. equal horizontal and vertical factors).
panX	Horizontal panning factor. By default, no panning occurs.

◆ DrawInspectionRegion() [5/5]

void eVision::EChecker::DrawInspectionRegion	(	eVision::EDrawAdapter *	graphicContext,
		float	zoomX,
		float	zoomY,
		float	panX,
		float	panY ) const

Draws the inspection region on EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage or EChecker::GetLastRegisteredImage.

Parameters

graphicContext	Device context of the drawing window.
zoomX	Magnification factor for zooming in or out in the horizontal direction. By default, the image is displayed in true scale.
zoomY	Magnification factor for zooming in or out in the vertical direction. Setting a 0 value (which is the default) will result in isotropic scaling (i.e. equal horizontal and vertical factors).
panX	Horizontal panning factor. By default, no panning occurs.
panY	Vertical panning factor. By default, no panning occurs.

◆ GetHighThresholdImage()

const eVision::EImageBW8 & eVision::EChecker::GetHighThresholdImage ( ) const

High threshold image.

◆ GetInspectionRegion()

const eVision::ERegion & eVision::EChecker::GetInspectionRegion ( ) const

Represents the inspection region, overlayed on EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage or EChecker::GetLastRegisteredImage.

◆ GetInspectionTolerance()

float eVision::EChecker::GetInspectionTolerance ( ) const

Inspection Tolerance factor. If you need to be more tolerant towards noise, texture or illumination irregularities, raise this value. If you have clean images, low texture and some defects are missed, lower this value. Default value: 4.0.

◆ GetIsTrained()

bool eVision::EChecker::GetIsTrained ( ) const

Application state. Has the checker been trained.

◆ GetLastRegisteredImage()

const eVision::EImageBW8 & eVision::EChecker::GetLastRegisteredImage ( ) const

Represents the last source image, after it has been aligned with the reference image and normalized.

◆ GetLowThresholdImage()

const eVision::EImageBW8 & eVision::EChecker::GetLowThresholdImage ( ) const

Low threshold image.

◆ GetNormalizationMode()

eVision::ENormalizationMode eVision::EChecker::GetNormalizationMode ( ) const

Normalization mode. This setting allows you to specify if you want to enable normalization, and if yes, which type. Normalization allows the training and inspection processes to automatically correct global illumination differences between the images. Default: Moments based normalization.

If you have consistent illumination between your images, using normalization can prove detrimental, as in that case the defect themselves might be enough to shift the global contrast.

◆ GetTrainingMode()

eVision::ETrainingMode eVision::EChecker::GetTrainingMode ( ) const

Training mode. The training mode determines which process will be used to build the threshold images used for inspection. EChecker has two training modes: A Quick mode, efficient for crude defects and stable images, and a Precise mode, designed for more subtle defects and handling more variability in the images. Default: Precise training mode.

◆ Inspect() [1/2]

void eVision::EChecker::Inspect	(	const eVision::EROIBW8 &	roi,
		const eVision::EFrame &	frame,
		eVision::ECodedImage2 &	defects )

Inspect the image. The defects are returned in the referential images aligned on the inspection region. See EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage and EChecker::GetLastRegisteredImage.

Frame is assumed to be a default frame when not specified.

Parameters

roi	ROI or Image to inspect.
frame	Frame relative to roi.
defects	List of defects returned by the inspection.

◆ Inspect() [2/2]

void eVision::EChecker::Inspect	(	const eVision::EROIBW8 &	roi,
		eVision::ECodedImage2 &	defects )

Inspect the image. The defects are returned in the referential images aligned on the inspection region. See EChecker::GetLowThresholdImage, EChecker::GetHighThresholdImage and EChecker::GetLastRegisteredImage.

Frame is assumed to be a default frame when not specified.

Parameters

roi	ROI or Image to inspect.
defects	List of defects returned by the inspection.

◆ Load() [1/2]

void eVision::EChecker::Load ( const std::string & file )

Load an inspection model.

Parameters

file File path

◆ Load() [2/2]

void eVision::EChecker::Load ( eVision::ESerializer * serializer )

Load an inspection model.

Parameters

serializer Serializer. Must be in read mode

◆ operator=()

eVision::EChecker & eVision::EChecker::operator= ( const eVision::EChecker & other )

Assignment operator.

Parameters

other The EChecker to copy into this one.

◆ ResetTraining()

void eVision::EChecker::ResetTraining ( )

Reset training. Use this if you want to reset the state of the object.

◆ Save() [1/2]

void eVision::EChecker::Save ( const std::string & file )

Save the inspection model.

Parameters

file File path

◆ Save() [2/2]

void eVision::EChecker::Save ( eVision::ESerializer * serializer )

Save the inspection model.

Parameters

serializer Serializer. Must be in write mode

◆ SetInspectionTolerance()

void eVision::EChecker::SetInspectionTolerance ( float tolerance )

Inspection Tolerance factor. If you need to be more tolerant towards noise, texture or illumination irregularities, raise this value. If you have clean images, low texture and some defects are missed, lower this value. Default value: 4.0.

◆ SetNormalizationMode()

void eVision::EChecker::SetNormalizationMode ( eVision::ENormalizationMode mode )

Normalization mode. This setting allows you to specify if you want to enable normalization, and if yes, which type. Normalization allows the training and inspection processes to automatically correct global illumination differences between the images. Default: Moments based normalization.

If you have consistent illumination between your images, using normalization can prove detrimental, as in that case the defect themselves might be enough to shift the global contrast.

◆ SetTrainingMode()

void eVision::EChecker::SetTrainingMode ( eVision::ETrainingMode mode )

Training mode. The training mode determines which process will be used to build the threshold images used for inspection. EChecker has two training modes: A Quick mode, efficient for crude defects and stable images, and a Precise mode, designed for more subtle defects and handling more variability in the images. Default: Precise training mode.

◆ Train() [1/12]

void eVision::EChecker::Train ( const std::vector< const eVision::EROIBW8 * > & rois )

Train. The distribution of noise is estimated for each pixel based on the training images. At inspection, a pixel is validated it if is close enough to the average value on the training set. The size of that interval is proportional to EChecker::GetInspectionTolerance.