EClassificationDataset manages a dataset of images. More...

#include <eVision/EasyDeepLearning/EClassificationDataset.h>

Inheritance diagram for eVision::EasyDeepLearning::EClassificationDataset:

Collaboration diagram for eVision::EasyDeepLearning::EClassificationDataset:

Public Member Functions
	EClassificationDataset ()
	Constructs a EClassificationDataset object.
	EClassificationDataset (const eVision::EasyDeepLearning::EClassificationDataset &other)
	Copy constructor of a EClassificationDataset object.
	EClassificationDataset (const eVision::EasyDeepLearning::EDataset &other)
	Copy constructor of a EClassificationDataset object from a EDataset object.
virtual	~EClassificationDataset () throw ()
eVision::EasyDeepLearning::EClassificationDataset &	operator= (const eVision::EasyDeepLearning::EClassificationDataset &other)
	Assignment operator.
virtual void	Clear ()
	Clears the dataset.
std::vector< eVision::EBaseROI * >	GetImages () const
	Gets a copy of all images in the dataset. If data augmentation is enabled, the returned images will be augmented versions of the ones in the dataset.
std::vector< eVision::EBaseROI * >	GetImages (const std::string &label) const
	Gets all the images associated with the given label.
std::unique_ptr< eVision::EBaseROI >	GetImageCopyWithDataAugmentation (int index, eVision::EasyDeepLearning::EDLDataAugmentationType generationType) const
	Generates a new image from the i-th image of the dataset. For the variant where a pointer is returned, the user is responsible for clearing the memory of the returned pointer.
std::unique_ptr< eVision::EBaseROI >	GetImageCopyWithDataAugmentation (int index) const
	Generates a new image from the i-th image of the dataset. For the variant where a pointer is returned, the user is responsible for clearing the memory of the returned pointer.
void	GetImageCopyWithDataAugmentation (int index, eVision::EBaseROI &img, eVision::EasyDeepLearning::EDLDataAugmentationType generationType)
	Generates a new image from the i-th image of the dataset. For the variant where a pointer is returned, the user is responsible for clearing the memory of the returned pointer.
void	GetImageCopyWithDataAugmentation (int index, eVision::EBaseROI &img)
	Generates a new image from the i-th image of the dataset. For the variant where a pointer is returned, the user is responsible for clearing the memory of the returned pointer.
void	SetEnableDataAugmentation (bool enable)
	Enable data augmentation.
bool	GetEnableDataAugmentation () const
	Enable data augmentation.
void	SetMaxHorizontalShift (int32_t maxShift)
	Maximum horizontal shift for data augmentation.
int32_t	GetMaxHorizontalShift () const
	Maximum horizontal shift for data augmentation.
void	SetMaxVerticalShift (int32_t maxShift)
	Maximum vertical shift for data augmentation.
int32_t	GetMaxVerticalShift () const
	Maximum vertical shift for data augmentation.
void	SetMaxRotationAngle (float maxAngle)
	Maximum rotation angle for data augmentation.
float	GetMaxRotationAngle () const
	Maximum rotation angle for data augmentation.
void	SetEnableVerticalFlip (bool enable)
	Enable vertical flipping in data augmentation.
bool	GetEnableVerticalFlip () const
	Enable vertical flipping in data augmentation.
void	SetEnableHorizontalFlip (bool enable)
	Enable horizontal flipping in data augmentation.
bool	GetEnableHorizontalFlip () const
	Enable horizontal flipping in data augmentation.
void	SetMinScale (float minScale)
	Minimum scaling allowed for data augmentation.
float	GetMinScale () const
	Minimum scaling allowed for data augmentation.
void	SetMaxScale (float maxScale)
	Maximum scaling allowed for data augmentation.
float	GetMaxScale () const
	Maximum scaling allowed for data augmentation.
void	SetMaxHorizontalShear (float hShear)
	Maximum absolute horizontal shear.
float	GetMaxHorizontalShear () const
	Maximum absolute horizontal shear.
void	SetMaxVerticalShear (float vShear)
	Maximum absolute vertical shear.
float	GetMaxVerticalShear () const
	Maximum absolute vertical shear.
void	SetMaxBrightnessOffset (float offset)
	Maximum absolute brightness offset. Its value must be between 0 and 1.
float	GetMaxBrightnessOffset () const
	Maximum absolute brightness offset. Its value must be between 0 and 1.
void	SetMaxContrastGain (float val)
	Maximum contrast gain. Its value must be strictly positive and over EClassificationDataset::SetMinContrastGain.
float	GetMaxContrastGain () const
	Maximum contrast gain. Its value must be strictly positive and over EClassificationDataset::SetMinContrastGain.
void	SetMinContrastGain (float val)
	Minimum contrast gain. Its value must be strictly positive and below EClassificationDataset::SetMaxContrastGain.
float	GetMinContrastGain () const
	Minimum contrast gain. Its value must be strictly positive and below EClassificationDataset::SetMaxContrastGain.
void	SetMinGamma (float gamma)
	Minimum gamma for gamma correction. Its value must be strictly positive and below EClassificationDataset::SetMaxGamma.
float	GetMinGamma () const
	Minimum gamma for gamma correction. Its value must be strictly positive and below EClassificationDataset::SetMaxGamma.
void	SetMaxGamma (float gamma)
	Maximum gamma for gamma correction. Its value must be higher than EClassificationDataset::SetMinGamma.
float	GetMaxGamma () const
	Maximum gamma for gamma correction. Its value must be higher than EClassificationDataset::SetMinGamma.
void	SetMaxHueOffset (float hueOffset)
	Maximum absolute hue offset. Its value must be between 0 and 180 degrees.
float	GetMaxHueOffset () const
	Maximum absolute hue offset. Its value must be between 0 and 180 degrees.
void	SetMinSaturationGain (float saturationGain)
	Minimum saturation gain. Its value must be strictly positive.
float	GetMinSaturationGain () const
	Minimum saturation gain. Its value must be strictly positive.
void	SetMaxSaturationGain (float saturationGain)
	Maximum saturation gain. Its value must be over or equal to EClassificationDataset::SetMinSaturationGain.
float	GetMaxSaturationGain () const
	Maximum saturation gain. Its value must be over or equal to EClassificationDataset::SetMinSaturationGain.
void	SetGaussianNoiseMaximumStandardDeviation (float gaussianMaximumDeviation)
	The Gaussian noise maximum standard deviation.
float	GetGaussianNoiseMaximumStandardDeviation () const
	The Gaussian noise maximum standard deviation.
void	SetGaussianNoiseMinimumStandardDeviation (float gaussianMinimumDeviation)
	The Gaussian noise minimum standard deviation.
float	GetGaussianNoiseMinimumStandardDeviation () const
	The Gaussian noise minimum standard deviation.
void	SetSpeckleNoiseMaximumStandardDeviation (float speckleMaximumDeviation)
	The speckle noise maximum standard deviation.
float	GetSpeckleNoiseMaximumStandardDeviation () const
	The speckle noise maximum standard deviation.
void	SetSpeckleNoiseMinimumStandardDeviation (float speckleMinimumDeviation)
	The speckle noise minimum standard deviation.
float	GetSpeckleNoiseMinimumStandardDeviation () const
	The speckle noise minimum standard deviation.
void	SetSaltAndPepperNoiseMaximumDensity (float saltAndPepperMaximumDensity)
	The maximum density of the salt and pepper noise.
float	GetSaltAndPepperNoiseMaximumDensity () const
	The maximum density of the salt and pepper noise.
void	SetSaltAndPepperNoiseMinimumDensity (float saltAndPepperMinimumDensity)
	The minimum density of the salt and pepper noise.
float	GetSaltAndPepperNoiseMinimumDensity () const
	The minimum density of the salt and pepper noise.
void	SplitDataset (eVision::EasyDeepLearning::EClassificationDataset &d1, eVision::EasyDeepLearning::EClassificationDataset &d2, float proportion, bool random) const
	Splits the dataset in two parts to be used for training and validation respectively.
void	SplitDataset (eVision::EasyDeepLearning::EClassificationDataset &d1, eVision::EasyDeepLearning::EClassificationDataset &d2, float proportion) const
	Splits the dataset in two parts to be used for training and validation respectively.
void	SplitDataset (eVision::EasyDeepLearning::EClassificationDataset &d1, eVision::EasyDeepLearning::EClassificationDataset &d2) const
	Splits the dataset in two parts to be used for training and validation respectively.
void	SplitDatasetForSegmentation (eVision::EasyDeepLearning::EClassificationDataset &d1, eVision::EasyDeepLearning::EClassificationDataset &d2, float proportion, bool random) const
	Splits the dataset in two parts for a supervised segmenter. The two parts are to be used for training and validation respectively.
void	SplitDatasetForSegmentation (eVision::EasyDeepLearning::EClassificationDataset &d1, eVision::EasyDeepLearning::EClassificationDataset &d2, float proportion) const
	Splits the dataset in two parts for a supervised segmenter. The two parts are to be used for training and validation respectively.
void	SplitDatasetForSegmentation (eVision::EasyDeepLearning::EClassificationDataset &d1, eVision::EasyDeepLearning::EClassificationDataset &d2) const
	Splits the dataset in two parts for a supervised segmenter. The two parts are to be used for training and validation respectively.
void	SplitDatasetForLocator (eVision::EasyDeepLearning::EClassificationDataset &d1, eVision::EasyDeepLearning::EClassificationDataset &d2, float proportion, bool random) const
	Splits the dataset in two parts for training and validation of a ELocator tool. Images without object labeling are excluded from the split.
void	SplitDatasetForLocator (eVision::EasyDeepLearning::EClassificationDataset &d1, eVision::EasyDeepLearning::EClassificationDataset &d2, float proportion) const
	Splits the dataset in two parts for training and validation of a ELocator tool. Images without object labeling are excluded from the split.
void	SplitDatasetForLocator (eVision::EasyDeepLearning::EClassificationDataset &d1, eVision::EasyDeepLearning::EClassificationDataset &d2) const
	Splits the dataset in two parts for training and validation of a ELocator tool. Images without object labeling are excluded from the split.
eVision::EasyDeepLearning::EClassificationDataset	ExtractSplit (const eVision::EasyDeepLearning::EDatasetSplit &split, eVision::EasyDeepLearning::EDatasetType type) const
	Creates a new EClassificationDataset containing only the images of the given split type.
	EClassificationDataset (CppWrapperInternal::InternalConstructor)
Public Member Functions inherited from eVision::EasyDeepLearning::EDataset
	EDataset ()
	Constructs a EDataset object.
	EDataset (const eVision::EasyDeepLearning::EDataset &other)
	Copy constructor of a EDataset object.
virtual	~EDataset () throw ()
eVision::EasyDeepLearning::EDataset &	operator= (const eVision::EasyDeepLearning::EDataset &other)
	Assignment operator.
void	Load (const std::string &path)
	Loads a classification dataset from disk.
void	Load (eVision::ESerializer *serializer)
	Loads a classification dataset from disk.
void	Save (const std::string &path)
	Saves a classification dataset to disk, containing the file paths to the images in the dataset and their associated label.
void	Save (eVision::ESerializer *serializer)
	Saves a classification dataset to disk, containing the file paths to the images in the dataset and their associated label.
void	SaveWithRelativePath (const std::string &path)
	Saves the dataset with image paths relative to the directory in which the dataset file is saved.
void	SetBasePath (const std::string &basePath)
	Sets the base path for the images specified with a relative path.
void	Export (const std::string &directory, eVision::EImageFileType fileType, int quality) const
	Exports the dataset and its images to the given directory.
void	Export (const std::string &directory, eVision::EImageFileType fileType) const
	Exports the dataset and its images to the given directory.
void	Export (const std::string &directory) const
	Exports the dataset and its images to the given directory.
void	Export (const std::string &directory, bool keepFilename, eVision::EImageFileType fileType, int quality) const
	Exports the dataset and its images to the given directory.
void	Export (const std::string &directory, bool keepFilename, eVision::EImageFileType fileType) const
	Exports the dataset and its images to the given directory.
void	Export (const std::string &directory, bool keepFilename) const
	Exports the dataset and its images to the given directory.
void	Export (const std::string &directory, int width, int height, int channels, eVision::EImageFileType fileType, int quality) const
	Exports the dataset and its images to the given directory.
void	Export (const std::string &directory, int width, int height, int channels, eVision::EImageFileType fileType) const
	Exports the dataset and its images to the given directory.
void	Export (const std::string &directory, int width, int height, int channels) const
	Exports the dataset and its images to the given directory.
void	Export (const std::string &directory, int width, int height, int channels, bool keepFilename, eVision::EImageFileType fileType, int quality) const
	Exports the dataset and its images to the given directory.
void	Export (const std::string &directory, int width, int height, int channels, bool keepFilename, eVision::EImageFileType fileType) const
	Exports the dataset and its images to the given directory.
void	Export (const std::string &directory, int width, int height, int channels, bool keepFilename) const
	Exports the dataset and its images to the given directory.
uint32_t	GetWidth () const
	Width of the region of interest of the first image added to the dataset.
uint32_t	GetHeight () const
	Height of the region of interest of the first image added to the dataset.
uint32_t	GetChannels () const
	Number of channels of the first image added to the dataset.
eVision::EasyDeepLearning::EImageAnnotationFormat	GetAvailableImageAnnotationFormat (const std::string &imageFile) const
	Image annotation file formats that are available next to the image file.
eVision::EasyDeepLearning::EImageAnnotationFormat	GetAvailableImageAnnotationFormat (int imgId) const
	Image annotation file formats that are available next to the image file.
void	ImportYOLOTXTAnnotations (int imgId, const std::string &labelFile)
	Imports YOLO TXT annotations (for EasyLocate Bounding Box).
void	ImportYOLOTXTAnnotations (const std::string &labelFile)
	Imports YOLO TXT annotations (for EasyLocate Bounding Box).
void	ImportPascalVOCXMLAnnotations (int imgId)
	Imports Pascal VOC XML annotations (for EasyLocate Bounding Box).
void	ImportPascalVOCXMLAnnotations ()
	Imports Pascal VOC XML annotations (for EasyLocate Bounding Box).
int	AddImage (const std::string &imagePath)
	Adds an image to the dataset.
int	AddImage (const std::string &imagePath, const std::string &label)
	Adds an image to the dataset.
int	AddImage (const std::string &imagePath, int originX, int originY, int width, int height, const eVision::ERegion &mask)
	Adds an image to the dataset.
int	AddImage (const std::string &imagePath, int originX, int originY, int width, int height, const eVision::ERegion &mask, const std::string &label)
	Adds an image to the dataset.
int	AddImage (const eVision::EBaseROI &img)
	Adds an image to the dataset.
int	AddImage (const eVision::EBaseROI &img, const std::string &label)
	Adds an image to the dataset.
int	AddImage (const eVision::EBaseROI &img, int originX, int originY, int width, int height, const eVision::ERegion &mask)
	Adds an image to the dataset.
int	AddImage (const eVision::EBaseROI &img, int originX, int originY, int width, int height, const eVision::ERegion &mask, const std::string &label)
	Adds an image to the dataset.
int	AddImages (const std::string &filter)
	Adds all the images present in the directory specified by the parameter path and whose filename matches the filter.
int	AddImages (const std::string &filter, const std::string &label)
	Adds all the images present in the directory specified by the parameter path and whose filename matches the filter.
int	AddImages (const std::string &filter, int originX, int originY, int width, int height, const eVision::ERegion &mask)
	Adds all the images present in the directory specified by the parameter path and whose filename matches the filter.
int	AddImages (const std::string &filter, int originX, int originY, int width, int height, const eVision::ERegion &mask, const std::string &label)
	Adds all the images present in the directory specified by the parameter path and whose filename matches the filter.
int	AddImagesFromCSV (const std::string &csvFile, int imageColumn, int labelColumn, int skipLines, const std::string &separator)
	Adds images using a CSV (Comma Separated Value) file where each row correspond to an image and the file path and label of each image is given in the columns imageColumn and labelColumn respectively.
int	AddImagesFromCSV (const std::string &csvFile, int imageColumn, int labelColumn, int skipLines)
	Adds images using a CSV (Comma Separated Value) file where each row correspond to an image and the file path and label of each image is given in the columns imageColumn and labelColumn respectively.
int	AddImagesFromCSV (const std::string &csvFile, int imageColumn, int labelColumn)
	Adds images using a CSV (Comma Separated Value) file where each row correspond to an image and the file path and label of each image is given in the columns imageColumn and labelColumn respectively.
int	AddImagesFromCSV (const std::string &csvFile, int imageColumn)
	Adds images using a CSV (Comma Separated Value) file where each row correspond to an image and the file path and label of each image is given in the columns imageColumn and labelColumn respectively.
int	AddImagesFromCSV (const std::string &csvFile)
	Adds images using a CSV (Comma Separated Value) file where each row correspond to an image and the file path and label of each image is given in the columns imageColumn and labelColumn respectively.
void	RemoveImage (int imgId)
	Removes the image at the given index. All annotations (label, segmentation, objects) will be lost.
int	GetNumImages () const
	Number of images in the dataset.
int	GetNumImageFiles () const
	Number of different image files contained in the dataset.
int	GetNumImagesForFile (const std::string &filePath) const
	Get the number of images in the dataset that are associated with the given image file.
std::unique_ptr< eVision::EBaseROI >	GetImageCopy (int index) const
	Gets a copy of the i-th image of the dataset.
void	GetImageCopy (int index, eVision::EBaseROI &img) const
	Gets a copy of the i-th image of the dataset.
std::string	GetImagePath (int index) const
	Gets the path of the i-th image of the dataset. If the image was not given using a path, the method will throw an exception.
bool	IsImageFile (int index) const
	Whether the image is stored as a file path towards an image.
bool	IsEmbeddedImage (int index) const
	Whether the image is embedded into the dataset and is not a reference towards an image file.
std::string	GetImageLabel (int index) const
	Gets the label of the i-th image of the dataset. An exception is thrown if the image has no classification label.
void	SetImageLabel (int index, const std::string &label)
	Sets the label of images in the dataset.
void	SetImageLabel (const std::string &filter, const std::string &label)
	Sets the label of images in the dataset.
void	UnsetImageLabel (int index)
	Unset the label of the given image of the dataset. After this operation, the given image will have no classification labels.
bool	HasLabel (int index) const
	Whether an image has a classification label.
int	GetNumLabeledImages () const
	Number of images that have a classification label.
int	GetNumUnlabeledImages () const
	Number of images that doesn't have any classification label.
std::vector< uint32_t >	GetImagesIndexesWithLabel (const std::string &label) const
	Gets a list of index corresponding to the images associated with the given label.
std::vector< uint32_t >	GetImagesIndexesWithLabel (int labelIndex) const
	Gets a list of index corresponding to the images associated with the given label.
std::vector< int >	GetLabeledImagesIndexes () const
	Gets a list of index corresponding to the images that have a classification label.
std::vector< int >	GetTextLabeledImagesIndexes () const
	Gets a list of index corresponding to the images that have a text only label.
std::vector< int >	GetImagesIndexesWithNoLabel () const
	Gets a list of index corresponding to the images with no classification labels.
int	GetNumLabels () const
	Number of labels in the dataset.
void	AddLabel (const std::string &label)
	Adds a label to the dataset.
std::string	GetLabel (int i) const
	Gets the i-th label of the dataset (starting from 0 to EDataset::GetNumLabels - 1).
void	SetLabel (int index, const std::string &label)
	Sets the i-th label of the dataset (starting from 0 to EDataset::GetNumLabels - 1). This operation does not add a new label to the dataset but simply renames an existing label.
float	GetLabelWeight (int index) const
	Gets the weight associated to the i-th label of the dataset (starting from 0 to EDataset::GetNumLabels - 1).
void	SetLabelWeight (int index, const float &weight)
	Sets the weight associated to the i-th label of the dataset (starting from 0 to EDataset::GetNumLabels - 1). This operation does not add a new label.
void	SetLabelColor (int index, const eVision::ERGBColor &color)
	Sets the label color for a classification label.
eVision::ERGBColor	GetLabelColor (int index) const
	Gets the label color for a classification label.
void	SetLabelOpacity (int index, const uint8_t opacity)
	Sets the label opacity for a classification label.
uint8_t	GetLabelOpacity (int index) const
	Gets the label opacity for a classification label.
void	RemoveLabel (const std::string &label)
	Removes a label from the dataset.
void	RemoveLabel (int labelId)
	Removes a label from the dataset.
void	SetSegmentationMap (int imageIndex, const eVision::EROIBW16 &segmentationMap)
	Sets the segmentation map of an image. The segmentation map is a 16-bit EROIBW16 image where the value of each pixel is equal to the corresponding segmentation label index (see EDataset::GetSegmentationLabel).
eVision::EImageBW16	GetSegmentationMap (int imageIndex) const
	Gets the segmentation map of an image. The segmentation map is a 16-bit EROIBW16 image where the value of each pixel is equal to the corresponding segmentation label index (see EDataset::GetSegmentationLabel).
void	UnsetSegmentation (int imageIndex)
	Unsets the segmentation of an image. After this operation, the image will have no segmentation.
void	ResetSegmentation (int imageIndex)
	Resets the segmentation of an image by setting all pixels to background.
void	AddRegionToSegment (int imageIndex, int segmentationLabelIndex, const eVision::ERegion &region)
	Adds a region (see ERegion) to the given segment of an image. If, before this call, the image had no segmentation, the pixels not in the given region will be set to the background segmentation label.
void	AddRegionToSegment (int imageIndex, const std::string &label, const eVision::ERegion &region)
	Adds a region (see ERegion) to the given segment of an image. If, before this call, the image had no segmentation, the pixels not in the given region will be set to the background segmentation label.
eVision::ERegion	GetRegionForSegment (int imageIndex, int segmentationLabelIndex)
	Gets a region corresponding to the pixels of the given segment.
eVision::ERegion	GetRegionForSegment (int imageIndex, const std::string &segmentationLabel)
	Gets a region corresponding to the pixels of the given segment.
int	GetNumSegmentationLabels () const
	Number of segmentation labels. A dataset has always at least one segmentation label: "background".
int	GetNumImagesForSegmentationLabel (int labelId) const
	Number of images containing the given segmentation label.
int	GetNumImagesForSegmentationLabel (const std::string &label) const
	Number of images containing the given segmentation label.
uint64_t	GetNumPixelsForSegmentationLabel (int labelId) const
	Number of pixels assigned to the given segmentation label.
uint64_t	GetNumPixelsForSegmentationLabel (const std::string &label) const
	Number of pixels assigned to the given segmentation label.
bool	HasSegmentation (int imageId) const
	Whether the image has a segmentation.
bool	HasForegroundSegments (int imageId) const
	Whether the image segmentation contains segments that are not background. If the image has no segmentation, this method throws an exception.
int	GetNumSegmentedBlobs (int imageId) const
	Number of segmented blobs in the image for all non-background labels or for a specific label.
int	GetNumSegmentedBlobs (int imageId, const std::string &label) const
	Number of segmented blobs in the image for all non-background labels or for a specific label.
int	GetNumSegmentedBlobs (int imageId, int labelId) const
	Number of segmented blobs in the image for all non-background labels or for a specific label.
std::string	GetSegmentationLabel (int index) const
	Gets the segmentation label.
void	SetSegmentationLabel (int index, const std::string &label)
	Sets the segmentation labels.
int	AddSegmentationLabel (const std::string &label, float labelWeight)
	Adds a segmentation label. The index of the new segmentation labels is returned by the method.
int	AddSegmentationLabel (const std::string &label)
	Adds a segmentation label. The index of the new segmentation labels is returned by the method.
void	RemoveSegmentationLabel (const std::string &label)
	Remove a segmentation label.
void	RemoveSegmentationLabel (int index)
	Remove a segmentation label.
float	GetSegmentationLabelWeight (int index) const
	Gets the segmentation label weights.
float	GetSegmentationLabelWeight (const std::string &label) const
	Gets the segmentation label weights.
void	SetSegmentationLabelWeight (int index, float weight)
	Sets the segmentation label weight.
void	SetSegmentationLabelWeight (const std::string &label, float weight)
	Sets the segmentation label weight.
void	SetSegmentationLabelColor (int index, const eVision::ERGBColor &color)
	Sets the label color for a segmentation label.
eVision::ERGBColor	GetSegmentationLabelColor (int index) const
	Gets the label color for a segmentation label.
void	SetSegmentationLabelOpacity (int index, const uint8_t opacity)
	Sets the label opacity for a segmentation label.
uint8_t	GetSegmentationLabelOpacity (int index) const
	Gets the label opacity for a segmentation label.
int	GetNumImagesWithSegmentation () const
	Number of images that have a ground truth segmentation.
int	GetNumImagesWithoutSegmentation () const
	Number of images that doesn't have a ground truth segmentation.
int	GetNumImagesWithoutForegroundSegments () const
	Number of images that have a ground truth segmentation that has no foreground segments and is thus entirely composed of background pixels.
int	GetNumImagesWithForegroundSegments () const
	Number of images that have a ground truth segmentation that has foreground segments and is this not entirely composed of background pixels.
void	ResetImageObjectLabeling (int imageIndex)
	Resets the object labeling for the specified image. This sets the image as having been labeled for object detection with no object present in the image.
void	UnsetImageObjectLabeling (int imageIndex)
	Unsets the object labeling for the specified image. This sets the image as not having been labeled for object detection. This image won't be use for training with a ELocator tool.
std::vector< eVision::EasyDeepLearning::ELocatorObject >	GetImageObjects (int imageIndex) const
	Gets all the objects of the specified image.
int	GetImageNumObjects (int imageIndex) const
	Number of objects for the specified image.
eVision::EasyDeepLearning::ELocatorObject	GetImageObject (int imageIndex, int objectIndex) const
	Gets the specified object for the given image.
void	SetImageObject (int imageIndex, int objectIndex, const eVision::EasyDeepLearning::ELocatorObject &obj)
	Sets the specified object for the given image.
void	SetImageObject (int imageIndex, int objectIndex, const std::string &label)
	Sets the specified object for the given image.
void	SetImageObject (int imageIndex, int objectIndex, const eVision::ERectangleRegion &region)
	Sets the specified object for the given image.
void	AddImageObject (int imageIndex, const eVision::EasyDeepLearning::ELocatorObject &obj)
	Adds an object to the image.
void	AddImageObject (int imageIndex, const std::string &label, const eVision::ERectangleRegion &box)
	Adds an object to the image.
void	RemoveImageObject (int imageIndex, int objectIndex)
	Removes an object from an image.
int	GetNumObjectLabels () const
	Number of object labels.
std::string	GetObjectLabel (int labelIndex) const
	Object label.
void	SetObjectLabel (int labelIndex, const std::string &newLabel)
	Sets an object label.
void	SetObjectLabel (const std::string &oldLabel, const std::string &newLabel)
	Sets an object label.
void	AddObjectLabel (const std::string &label, float weight)
	Adds an object label.
void	AddObjectLabel (const std::string &label)
	Adds an object label.
void	RemoveObjectLabel (int labelIndex)
	Removes an object label.
void	RemoveObjectLabel (const std::string &label)
	Removes an object label.
void	SetObjectLabelWeight (int labelIndex, float weight)
	Sets the weight of an object label.
void	SetObjectLabelWeight (const std::string &label, float weight)
	Sets the weight of an object label.
float	GetObjectLabelWeight (int labelIndex) const
	Gets the weight of an object label.
float	GetObjectLabelWeight (const std::string &label) const
	Gets the weight of an object label.
void	SetObjectLabelColor (int index, const eVision::ERGBColor &color)
	Sets the label color for an object label.
eVision::ERGBColor	GetObjectLabelColor (int index) const
	Gets the label color for an object label.
void	SetObjectLabelOpacity (int index, const uint8_t opacity)
	Sets the label opacity for an object label.
uint8_t	GetObjectLabelOpacity (int index) const
	Gets the label opacity for an object label.
bool	HasObjectLabeling (int imageIndex) const
	Whether the image is labeled for object detection and use with ELocator.
int	GetNumImagesWithObjectLabeling () const
	Number of images in the dataset that are labeled for object detection.
int	GetNumImagesWithoutObjectLabeling () const
	Number of images in the dataset that are not labeled for object detection.
int	GetNumImagesWithoutObjects () const
	Number of images in the dataset that are labeled for object detection but have been associated with no object.
int	GetNumImagesWithObjects () const
	Number of images in the dataset that are labeled for object detection and have 1 or more objects.
int	GetNumImagesWithObjectLabel (int labelIdx) const
	Number of images that contain an object with the given label.
int	GetNumImagesWithObjectLabel (const std::string &label) const
	Number of images that contain an object with the given label.
uint64_t	GetNumObjectsWithLabel (int labelIdx) const
	Number of objects with the given label.
uint64_t	GetNumObjectsWithLabel (const std::string &label) const
	Number of objects with the given label.
float	GetSameLabelMaxOverlap () const
	Maximum overlap between objects with the same label in the dataset.
float	GetAbsoluteMaxOverlap () const
	Absolute maximum overlap between objects in the dataset.
int	GetMaxNumObjectPerImage () const
	Maximum number of objects for an image in the dataset. If no images has object labeling (see EDataset::HasObjectLabeling), the method returns -1.
void	SetObjectSize (int objectSize)
	Object size for EasyLocate Interest point.
int	GetObjectSize () const
	Object size for EasyLocate Interest point.
bool	HasOCRAnnotations (int imgId) const
	Whether the specified image has OCR annotations.
void	UnsetOCRAnnotations (int imgId)
	Unsets the OCR labeling for the specified image.
std::vector< eVision::EasyDeepOCR::ETextLine >	GetOCRAnnotations (int imgId) const
	Ground truth text lines for the specified image.
void	SetOCRAnnotations (int imgId, const std::vector< eVision::EasyDeepOCR::ETextLine > &texts)
	Sets the ground truth text lines for the specified image.
int	GetNumImagesWithOCR () const
	Number of images with OCR annotations.
int	GetNumImagesWithoutOCR () const
	Number of images without OCR annotations.
int	GetNumTexts () const
	Number of annotated texts in the dataset.
eVision::EasyDeepLearning::EDataset	GetTextDataset () const
	Transforms the current dataset to have only subwindow containing the actual text to read.
eVision::EasyDeepLearning::EDataset	GetTextOrientationDataset () const
	Transforms the current text dataset to duplicate it with half image flipped, adding corresponding ground truth.
void	SplitDataset (eVision::EasyDeepLearning::EDataset &d1, eVision::EasyDeepLearning::EDataset &d2, float proportion, bool random) const
	Splits the dataset in two parts to be used for training and validation respectively.
void	SplitDataset (eVision::EasyDeepLearning::EDataset &d1, eVision::EasyDeepLearning::EDataset &d2, float proportion) const
	Splits the dataset in two parts to be used for training and validation respectively.
void	SplitDataset (eVision::EasyDeepLearning::EDataset &d1, eVision::EasyDeepLearning::EDataset &d2) const
	Splits the dataset in two parts to be used for training and validation respectively.
void	SplitDatasetForSegmentation (eVision::EasyDeepLearning::EDataset &d1, eVision::EasyDeepLearning::EDataset &d2, float proportion, bool random) const
	Splits the dataset in two parts for a supervised segmenter. The two parts are to be used for training and validation respectively.
void	SplitDatasetForSegmentation (eVision::EasyDeepLearning::EDataset &d1, eVision::EasyDeepLearning::EDataset &d2, float proportion) const
	Splits the dataset in two parts for a supervised segmenter. The two parts are to be used for training and validation respectively.
void	SplitDatasetForSegmentation (eVision::EasyDeepLearning::EDataset &d1, eVision::EasyDeepLearning::EDataset &d2) const
	Splits the dataset in two parts for a supervised segmenter. The two parts are to be used for training and validation respectively.
void	SplitDatasetForLocator (eVision::EasyDeepLearning::EDataset &d1, eVision::EasyDeepLearning::EDataset &d2, float proportion, bool random) const
	Splits the dataset in two parts for training and validation of a ELocator tool. Images without object labeling are excluded from the split.
void	SplitDatasetForLocator (eVision::EasyDeepLearning::EDataset &d1, eVision::EasyDeepLearning::EDataset &d2, float proportion) const
	Splits the dataset in two parts for training and validation of a ELocator tool. Images without object labeling are excluded from the split.
void	SplitDatasetForLocator (eVision::EasyDeepLearning::EDataset &d1, eVision::EasyDeepLearning::EDataset &d2) const
	Splits the dataset in two parts for training and validation of a ELocator tool. Images without object labeling are excluded from the split.
eVision::EasyDeepLearning::EDatasetSplit	GetSplit (float trainingProportion, float validationPropotion, eVision::EasyDeepLearning::EDeepLearningToolType toolType, const std::string &goodLabel) const
	Generates a split.
eVision::EasyDeepLearning::EDatasetSplit	GetSplit (float trainingProportion, float validationPropotion, eVision::EasyDeepLearning::EDeepLearningToolType toolType, const std::string &goodLabel, uint32_t seed) const
	Generates a split.
eVision::EasyDeepLearning::EDatasetSplit	GetSplit (int numTrainingImages, int numValidationImages, eVision::EasyDeepLearning::EDeepLearningToolType toolType, const std::string &goodLabel) const
	Generates a split.
eVision::EasyDeepLearning::EDatasetSplit	GetSplit (int numTrainingImages, int numValidationImages, eVision::EasyDeepLearning::EDeepLearningToolType toolType, const std::string &goodLabel, uint32_t seed) const
	Generates a split.
eVision::EasyDeepLearning::EDataset	ExtractSplit (const eVision::EasyDeepLearning::EDatasetSplit &split, eVision::EasyDeepLearning::EDatasetType type) const
	Creates a new EDataset containing only the images of the given split type.
void	SetRegionOfInterest (int imageIndex, int xOrg, int yOrg, int width, int height)
	Sets the region of interest for the specified image.
int	GetRegionOfInterestOriginX (int imageIndex) const
	Region of interest origin abscissa for the specified image.
int	GetRegionOfInterestOriginY (int imageIndex) const
	Region of interest origin ordinate for the specified image.
int	GetRegionOfInterestWidth (int imageIndex) const
	Region of interest width for the specified image.
int	GetRegionOfInterestHeight (int imageIndex) const
	Region of interest height for the specified image.
eVision::ERegion	GetMask (int imageIndex) const
	Gets the mask region/don't care area for the given image. The mask is defined with respect to the region of interest of the image.
void	SetMask (int imageIndex, const eVision::ERegion &mask)
	Sets the mask region/don't care area for the given image. The mask is defined with respect to the region of interest of the image.
	EDataset (CppWrapperInternal::InternalConstructor)

Additional Inherited Members
Public Attributes inherited from eVision::EasyDeepLearning::EDataset
CppWrapperInternal::WrappedClass	wrappedClass_

Detailed Description

EClassificationDataset manages a dataset of images.

A dataset is a collection of images with different types of labeling: labeling for the classification of images, labeling for the segmentation of pixels, and/or labeling for the detection of objects (classification and localization).

The dataset maintains 3 sets of labels for each type of labeling:

the classification labels that characterize an entire image;
the segmentation labels that characterize the pixels of an image; and
the object labels that characterize axis-aligned rectangle region of an image.

The classification and segmentation labels are entirely user defined. The set of segmentation labels will always contain at least the "Background" label representing pixels of the images that have no relevant information for your task (for example, in a defect segmentation application, the "Background" pixels would be the pixels without any defects).

For each type of labeling, an image can either be unlabeled or labeled. When an image in unlabeled for a given type of labeling, the image won't be used for training a deep learning tool that requires this type of labeling.

The image in the dataset can be stored as path to an image file or as an eVision image structure. Supported structures are 8-bits monochrome (EImageBW8), 16-bits monochrome (EImageBW16), and 24-bits color (EROIC24, EImageC24).

The dataset associates with each image a region of interest and a mask/don't care area. By default, the region of interest of an image is its full extent and its mask is empty.

A EClassificationDataset object is also responsible for providing tools to do data augmentation. Data augmentation is the process of generating new images on-the-fly by applying affine transformations to those already in the dataset. Data augmentation allows a deep neural network to be invariant to the applied transformation without having to capture and label real world images containing those transformations.

A dataset can contain images with different sizes (with and height of their region of interest). However, the dataset has a default resolution (see EDataset::GetWidth and EDataset::GetHeight) that is used by deep learning tools that require the same input image size such as the EClassifier. When the images have different sizes, the default resolution will be the resolution of the region of interest of the first image added to the dataset.

License(s): BasicTypes

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

Constructor & Destructor Documentation

◆ EClassificationDataset() [1/4]

eVision::EasyDeepLearning::EClassificationDataset::EClassificationDataset ( )

Constructs a EClassificationDataset object.

◆ EClassificationDataset() [2/4]

eVision::EasyDeepLearning::EClassificationDataset::EClassificationDataset ( const eVision::EasyDeepLearning::EClassificationDataset & other )

Copy constructor of a EClassificationDataset object.

Parameters

other Reference to the EClassificationDataset object that should be copied

◆ EClassificationDataset() [3/4]

eVision::EasyDeepLearning::EClassificationDataset::EClassificationDataset ( const eVision::EasyDeepLearning::EDataset & other )

Copy constructor of a EClassificationDataset object from a EDataset object.

Parameters

other Reference to the EDataset object that should be copied

◆ ~EClassificationDataset()

eVision::EasyDeepLearning::EClassificationDataset::~EClassificationDataset	(		)
throw	(	)

virtual

◆ EClassificationDataset() [4/4]

eVision::EasyDeepLearning::EClassificationDataset::EClassificationDataset ( CppWrapperInternal::InternalConstructor )

Member Function Documentation

◆ Clear()

void eVision::EasyDeepLearning::EClassificationDataset::Clear ( )

virtual

Clears the dataset.

Reimplemented from eVision::EasyDeepLearning::EDataset.

◆ ExtractSplit()

eVision::EasyDeepLearning::EClassificationDataset eVision::EasyDeepLearning::EClassificationDataset::ExtractSplit	(	const eVision::EasyDeepLearning::EDatasetSplit &	split,
		eVision::EasyDeepLearning::EDatasetType	type ) const

Creates a new EClassificationDataset containing only the images of the given split type.

Parameters

split	Split
type	Type of the split to extract

◆ GetEnableDataAugmentation()

bool eVision::EasyDeepLearning::EClassificationDataset::GetEnableDataAugmentation ( ) const

Enable data augmentation.

◆ GetEnableHorizontalFlip()

bool eVision::EasyDeepLearning::EClassificationDataset::GetEnableHorizontalFlip ( ) const

Enable horizontal flipping in data augmentation.

◆ GetEnableVerticalFlip()

bool eVision::EasyDeepLearning::EClassificationDataset::GetEnableVerticalFlip ( ) const

Enable vertical flipping in data augmentation.

◆ GetGaussianNoiseMaximumStandardDeviation()

float eVision::EasyDeepLearning::EClassificationDataset::GetGaussianNoiseMaximumStandardDeviation ( ) const

The Gaussian noise maximum standard deviation.

The Gaussian noise is an additive noise sampled from a Gaussian distribution of deviation between EClassificationDataset::SetGaussianNoiseMinimumStandardDeviation and EClassificationDataset::SetGaussianNoiseMaximumStandardDeviation (also called the normal distribution).

Its value must be superior or equal to EClassificationDataset::SetGaussianNoiseMinimumStandardDeviation.

This noise is computed before the salt and paper noise.

◆ GetGaussianNoiseMinimumStandardDeviation()

float eVision::EasyDeepLearning::EClassificationDataset::GetGaussianNoiseMinimumStandardDeviation ( ) const

The Gaussian noise minimum standard deviation.

The Gaussian noise is an additive noise sampled from a Gaussian distribution of deviation between EClassificationDataset::SetGaussianNoiseMinimumStandardDeviation and EClassificationDataset::SetGaussianNoiseMaximumStandardDeviation (also called the normal distribution).

Its value must be between 0 and EClassificationDataset::SetGaussianNoiseMaximumStandardDeviation.

This noise is computed before the salt and paper noise.

◆ GetImageCopyWithDataAugmentation() [1/4]

std::unique_ptr< eVision::EBaseROI > eVision::EasyDeepLearning::EClassificationDataset::GetImageCopyWithDataAugmentation ( int index ) const

Generates a new image from the i-th image of the dataset. For the variant where a pointer is returned, the user is responsible for clearing the memory of the returned pointer.

Parameters

index The index of the image.

If the data augmentation fails, the method will throw an exception.

◆ GetImageCopyWithDataAugmentation() [2/4]

std::unique_ptr< eVision::EBaseROI > eVision::EasyDeepLearning::EClassificationDataset::GetImageCopyWithDataAugmentation	(	int	index,
		eVision::EasyDeepLearning::EDLDataAugmentationType	generationType ) const

Generates a new image from the i-th image of the dataset. For the variant where a pointer is returned, the user is responsible for clearing the memory of the returned pointer.

Parameters

index	The index of the image.
generationType	The type of transformation to generate (default: random).

If the data augmentation fails, the method will throw an exception.

◆ GetImageCopyWithDataAugmentation() [3/4]

void eVision::EasyDeepLearning::EClassificationDataset::GetImageCopyWithDataAugmentation	(	int	index,
		eVision::EBaseROI &	img )

Generates a new image from the i-th image of the dataset. For the variant where a pointer is returned, the user is responsible for clearing the memory of the returned pointer.

Parameters

index	The index of the image.
img	Image object to copy the image into.

If the data augmentation fails, the method will throw an exception.

◆ GetImageCopyWithDataAugmentation() [4/4]

void eVision::EasyDeepLearning::EClassificationDataset::GetImageCopyWithDataAugmentation	(	int	index,
		eVision::EBaseROI &	img,
		eVision::EasyDeepLearning::EDLDataAugmentationType	generationType )

Generates a new image from the i-th image of the dataset. For the variant where a pointer is returned, the user is responsible for clearing the memory of the returned pointer.

Parameters

index	The index of the image.
img	Image object to copy the image into.
generationType	The type of transformation to generate (default: random).

If the data augmentation fails, the method will throw an exception.

◆ GetImages() [1/2]

std::vector< eVision::EBaseROI * > eVision::EasyDeepLearning::EClassificationDataset::GetImages ( ) const

Gets a copy of all images in the dataset. If data augmentation is enabled, the returned images will be augmented versions of the ones in the dataset.

The caller is responsible for clearing the memory allocated for each image.

The method will load images from their given path when necessary.

Deprecated: Deprecated in version 24.6. Scheduled to be removed in version 27.6.

◆ GetImages() [2/2]

std::vector< eVision::EBaseROI * > eVision::EasyDeepLearning::EClassificationDataset::GetImages ( const std::string & label ) const

Gets all the images associated with the given label.

If the parameter is empty, all the images are returned. If data augmentation is enabled, the returned images will be augmented versions of the ones in the dataset.

The caller is responsible for clearing the memory allocated for each image.

Parameters

label The label.

The method will load images from their given path when necessary.

Deprecated: Deprecated in version 24.6. Scheduled to be removed in version 27.6.

◆ GetMaxBrightnessOffset()

float eVision::EasyDeepLearning::EClassificationDataset::GetMaxBrightnessOffset ( ) const

Maximum absolute brightness offset. Its value must be between 0 and 1.

Brightness transformation is performed by adding a value taken between -EClassificationDataset::SetMaxBrightnessOffset and +EClassificationDataset::SetMaxBrightnessOffset to each pixel of the normalized image.

◆ GetMaxContrastGain()

float eVision::EasyDeepLearning::EClassificationDataset::GetMaxContrastGain ( ) const

Maximum contrast gain. Its value must be strictly positive and over EClassificationDataset::SetMinContrastGain.

Contrast transformation is performed by multiplying each mean-centered pixel value by a gain value taken between EClassificationDataset::SetMinContrastGain and EClassificationDataset::SetMaxContrastGain. A contrast transformation does not change the overall brightness of an image.

◆ GetMaxGamma()

float eVision::EasyDeepLearning::EClassificationDataset::GetMaxGamma ( ) const

Maximum gamma for gamma correction. Its value must be higher than EClassificationDataset::SetMinGamma.

Gamma correction transformation is performed by raising each normalized pixel value to the power of gamma with gamma taken between EClassificationDataset::SetMinGamma and EClassificationDataset::SetMaxGamma.

◆ GetMaxHorizontalShear()

float eVision::EasyDeepLearning::EClassificationDataset::GetMaxHorizontalShear ( ) const

Maximum absolute horizontal shear.

It is represented as an angle from the vertical direction. Its value must be between 0 and 90 degrees.

◆ GetMaxHorizontalShift()

int32_t eVision::EasyDeepLearning::EClassificationDataset::GetMaxHorizontalShift ( ) const

Maximum horizontal shift for data augmentation.

The horizontal shift will be between -EClassificationDataset::SetMaxHorizontalShift and +EClassificationDataset::SetMaxHorizontalShift.

◆ GetMaxHueOffset()

float eVision::EasyDeepLearning::EClassificationDataset::GetMaxHueOffset ( ) const

Maximum absolute hue offset. Its value must be between 0 and 180 degrees.

The hue is represented as an angle between 0 and 360 degrees. The hue transformation is performed by rotating the hue of each pixel by a value between -EClassificationDataset::SetMaxHueOffset and +EClassificationDataset::SetMaxHueOffset. This transformation only works for color images.

◆ GetMaxRotationAngle()

float eVision::EasyDeepLearning::EClassificationDataset::GetMaxRotationAngle ( ) const

Maximum rotation angle for data augmentation.

The rotation angle will be between -EClassificationDataset::SetMaxRotationAngle and +EClassificationDataset::SetMaxRotationAngle.

◆ GetMaxSaturationGain()

float eVision::EasyDeepLearning::EClassificationDataset::GetMaxSaturationGain ( ) const

Maximum saturation gain. Its value must be over or equal to EClassificationDataset::SetMinSaturationGain.

The saturation transformation is performed by multiplying the saturation of each pixel by a value between EClassificationDataset::SetMinSaturationGain and EClassificationDataset::SetMaxSaturationGain.

◆ GetMaxScale()

float eVision::EasyDeepLearning::EClassificationDataset::GetMaxScale ( ) const

Maximum scaling allowed for data augmentation.

◆ GetMaxVerticalShear()

float eVision::EasyDeepLearning::EClassificationDataset::GetMaxVerticalShear ( ) const

Maximum absolute vertical shear.

It is represented as an angle from the horizontal direction. Its value must be between 0 and 90 degrees.

◆ GetMaxVerticalShift()

int32_t eVision::EasyDeepLearning::EClassificationDataset::GetMaxVerticalShift ( ) const

Maximum vertical shift for data augmentation.

The vertical shift will be between -EClassificationDataset::SetMaxHorizontalShift and +EClassificationDataset::SetMaxHorizontalShift.

◆ GetMinContrastGain()

float eVision::EasyDeepLearning::EClassificationDataset::GetMinContrastGain ( ) const

Minimum contrast gain. Its value must be strictly positive and below EClassificationDataset::SetMaxContrastGain.

Contrast transformation is performed by multiplying each mean-centered pixel value by a gain value taken between EClassificationDataset::SetMinContrastGain and EClassificationDataset::SetMaxContrastGain. A contrast transformation does not change the overall brightness of an image.

◆ GetMinGamma()

float eVision::EasyDeepLearning::EClassificationDataset::GetMinGamma ( ) const

Minimum gamma for gamma correction. Its value must be strictly positive and below EClassificationDataset::SetMaxGamma.

Gamma correction transformation is performed by raising each normalized pixel value to the power of gamma with gamma taken between EClassificationDataset::SetMinGamma and

◆ GetMinSaturationGain()

float eVision::EasyDeepLearning::EClassificationDataset::GetMinSaturationGain ( ) const

Minimum saturation gain. Its value must be strictly positive.

The saturation transformation is performed by multiplying the saturation of each pixel by a value between EClassificationDataset::SetMinSaturationGain and EClassificationDataset::SetMaxSaturationGain.

◆ GetMinScale()

float eVision::EasyDeepLearning::EClassificationDataset::GetMinScale ( ) const

Minimum scaling allowed for data augmentation.

◆ GetSaltAndPepperNoiseMaximumDensity()

float eVision::EasyDeepLearning::EClassificationDataset::GetSaltAndPepperNoiseMaximumDensity ( ) const

The maximum density of the salt and pepper noise.

The salt and pepper noise sets the value of a number of randomly selected (between EClassificationDataset::SetSaltAndPepperNoiseMinimumDensity and EClassificationDataset::SetSaltAndPepperNoiseMaximumDensity) pixels to its minimum or maximum value.

Its value must be between EClassificationDataset::SetSaltAndPepperNoiseMinimumDensity and 1.

This noise is computed after all the other noises.

◆ GetSaltAndPepperNoiseMinimumDensity()

float eVision::EasyDeepLearning::EClassificationDataset::GetSaltAndPepperNoiseMinimumDensity ( ) const

The minimum density of the salt and pepper noise.

The salt and pepper noise sets the value of a number of randomly selected (between EClassificationDataset::SetSaltAndPepperNoiseMinimumDensity and EClassificationDataset::SetSaltAndPepperNoiseMaximumDensity) pixels to its minimum or maximum value.

Its value must be between 0 and EClassificationDataset::SetSaltAndPepperNoiseMaximumDensity.

This noise is computed after all the other noises.

◆ GetSpeckleNoiseMaximumStandardDeviation()

float eVision::EasyDeepLearning::EClassificationDataset::GetSpeckleNoiseMaximumStandardDeviation ( ) const

The speckle noise maximum standard deviation.

The speckle noise is a multiplicative noise sampled from a Gamma distribution with a mean of 1 and with its standard deviation between EClassificationDataset::SetSpeckleNoiseMinimumStandardDeviation and EClassificationDataset::SetSpeckleNoiseMaximumStandardDeviation.

Its value must be strictly higher than EClassificationDataset::SetSpeckleNoiseMinimumStandardDeviation.

This noise is computed before the salt and paper noise.

◆ GetSpeckleNoiseMinimumStandardDeviation()

float eVision::EasyDeepLearning::EClassificationDataset::GetSpeckleNoiseMinimumStandardDeviation ( ) const

The speckle noise minimum standard deviation.

The speckle noise is a multiplicative noise sampled from a Gamma distribution with a mean of 1 and with its standard deviation between EClassificationDataset::SetSpeckleNoiseMinimumStandardDeviation and EClassificationDataset::SetSpeckleNoiseMaximumStandardDeviation.

Its value must be strictly positive and lower than EClassificationDataset::SetSpeckleNoiseMaximumStandardDeviation.

This noise is computed before the salt and paper noise.

◆ operator=()

eVision::EasyDeepLearning::EClassificationDataset & eVision::EasyDeepLearning::EClassificationDataset::operator= ( const eVision::EasyDeepLearning::EClassificationDataset & other )

Assignment operator.

Parameters

other Reference to the EClassificationDataset object used for the assignment

◆ SetEnableDataAugmentation()

void eVision::EasyDeepLearning::EClassificationDataset::SetEnableDataAugmentation ( bool enable )

Enable data augmentation.

◆ SetEnableHorizontalFlip()

void eVision::EasyDeepLearning::EClassificationDataset::SetEnableHorizontalFlip ( bool enable )

Enable horizontal flipping in data augmentation.

◆ SetEnableVerticalFlip()

void eVision::EasyDeepLearning::EClassificationDataset::SetEnableVerticalFlip ( bool enable )

Enable vertical flipping in data augmentation.

◆ SetGaussianNoiseMaximumStandardDeviation()

void eVision::EasyDeepLearning::EClassificationDataset::SetGaussianNoiseMaximumStandardDeviation ( float gaussianMaximumDeviation )

The Gaussian noise maximum standard deviation.

The Gaussian noise is an additive noise sampled from a Gaussian distribution of deviation between EClassificationDataset::SetGaussianNoiseMinimumStandardDeviation and EClassificationDataset::SetGaussianNoiseMaximumStandardDeviation (also called the normal distribution).

Its value must be superior or equal to EClassificationDataset::SetGaussianNoiseMinimumStandardDeviation.

Parameters

gaussianMaximumDeviation The maximum standard deviation of the gaussian distribution expressed in percentage of the image type max intensity

This noise is computed before the salt and paper noise.

◆ SetGaussianNoiseMinimumStandardDeviation()

void eVision::EasyDeepLearning::EClassificationDataset::SetGaussianNoiseMinimumStandardDeviation ( float gaussianMinimumDeviation )

The Gaussian noise minimum standard deviation.

The Gaussian noise is an additive noise sampled from a Gaussian distribution of deviation between EClassificationDataset::SetGaussianNoiseMinimumStandardDeviation and EClassificationDataset::SetGaussianNoiseMaximumStandardDeviation (also called the normal distribution).

Its value must be between 0 and EClassificationDataset::SetGaussianNoiseMaximumStandardDeviation.

Parameters

gaussianMinimumDeviation The minimum standard deviation of the gaussian distribution expressed in percentage of the image type max intensity

This noise is computed before the salt and paper noise.

◆ SetMaxBrightnessOffset()

void eVision::EasyDeepLearning::EClassificationDataset::SetMaxBrightnessOffset ( float offset )

Maximum absolute brightness offset. Its value must be between 0 and 1.

Brightness transformation is performed by adding a value taken between -EClassificationDataset::SetMaxBrightnessOffset and +EClassificationDataset::SetMaxBrightnessOffset to each pixel of the normalized image.

◆ SetMaxContrastGain()

void eVision::EasyDeepLearning::EClassificationDataset::SetMaxContrastGain ( float val )

Maximum contrast gain. Its value must be strictly positive and over EClassificationDataset::SetMinContrastGain.

Contrast transformation is performed by multiplying each mean-centered pixel value by a gain value taken between EClassificationDataset::SetMinContrastGain and EClassificationDataset::SetMaxContrastGain. A contrast transformation does not change the overall brightness of an image.

◆ SetMaxGamma()

void eVision::EasyDeepLearning::EClassificationDataset::SetMaxGamma ( float gamma )

Maximum gamma for gamma correction. Its value must be higher than EClassificationDataset::SetMinGamma.

Gamma correction transformation is performed by raising each normalized pixel value to the power of gamma with gamma taken between EClassificationDataset::SetMinGamma and EClassificationDataset::SetMaxGamma.

◆ SetMaxHorizontalShear()

void eVision::EasyDeepLearning::EClassificationDataset::SetMaxHorizontalShear ( float hShear )

Maximum absolute horizontal shear.

It is represented as an angle from the vertical direction. Its value must be between 0 and 90 degrees.

◆ SetMaxHorizontalShift()

void eVision::EasyDeepLearning::EClassificationDataset::SetMaxHorizontalShift ( int32_t maxShift )

Maximum horizontal shift for data augmentation.

The horizontal shift will be between -EClassificationDataset::SetMaxHorizontalShift and +EClassificationDataset::SetMaxHorizontalShift.

◆ SetMaxHueOffset()

void eVision::EasyDeepLearning::EClassificationDataset::SetMaxHueOffset ( float hueOffset )

Maximum absolute hue offset. Its value must be between 0 and 180 degrees.

The hue is represented as an angle between 0 and 360 degrees. The hue transformation is performed by rotating the hue of each pixel by a value between -EClassificationDataset::SetMaxHueOffset and +EClassificationDataset::SetMaxHueOffset. This transformation only works for color images.

◆ SetMaxRotationAngle()

void eVision::EasyDeepLearning::EClassificationDataset::SetMaxRotationAngle ( float maxAngle )

Maximum rotation angle for data augmentation.

The rotation angle will be between -EClassificationDataset::SetMaxRotationAngle and +EClassificationDataset::SetMaxRotationAngle.

◆ SetMaxSaturationGain()

void eVision::EasyDeepLearning::EClassificationDataset::SetMaxSaturationGain ( float saturationGain )

Maximum saturation gain. Its value must be over or equal to EClassificationDataset::SetMinSaturationGain.

The saturation transformation is performed by multiplying the saturation of each pixel by a value between EClassificationDataset::SetMinSaturationGain and EClassificationDataset::SetMaxSaturationGain.

◆ SetMaxScale()

void eVision::EasyDeepLearning::EClassificationDataset::SetMaxScale ( float maxScale )

Maximum scaling allowed for data augmentation.

◆ SetMaxVerticalShear()

void eVision::EasyDeepLearning::EClassificationDataset::SetMaxVerticalShear ( float vShear )

Maximum absolute vertical shear.

It is represented as an angle from the horizontal direction. Its value must be between 0 and 90 degrees.

◆ SetMaxVerticalShift()

void eVision::EasyDeepLearning::EClassificationDataset::SetMaxVerticalShift ( int32_t maxShift )

Maximum vertical shift for data augmentation.

The vertical shift will be between -EClassificationDataset::SetMaxHorizontalShift and +EClassificationDataset::SetMaxHorizontalShift.

◆ SetMinContrastGain()

void eVision::EasyDeepLearning::EClassificationDataset::SetMinContrastGain ( float val )

Minimum contrast gain. Its value must be strictly positive and below EClassificationDataset::SetMaxContrastGain.

Contrast transformation is performed by multiplying each mean-centered pixel value by a gain value taken between EClassificationDataset::SetMinContrastGain and EClassificationDataset::SetMaxContrastGain. A contrast transformation does not change the overall brightness of an image.

◆ SetMinGamma()

void eVision::EasyDeepLearning::EClassificationDataset::SetMinGamma ( float gamma )

Minimum gamma for gamma correction. Its value must be strictly positive and below EClassificationDataset::SetMaxGamma.

Gamma correction transformation is performed by raising each normalized pixel value to the power of gamma with gamma taken between EClassificationDataset::SetMinGamma and

◆ SetMinSaturationGain()

void eVision::EasyDeepLearning::EClassificationDataset::SetMinSaturationGain ( float saturationGain )

Minimum saturation gain. Its value must be strictly positive.

The saturation transformation is performed by multiplying the saturation of each pixel by a value between EClassificationDataset::SetMinSaturationGain and EClassificationDataset::SetMaxSaturationGain.

◆ SetMinScale()

void eVision::EasyDeepLearning::EClassificationDataset::SetMinScale ( float minScale )

Minimum scaling allowed for data augmentation.

◆ SetSaltAndPepperNoiseMaximumDensity()

void eVision::EasyDeepLearning::EClassificationDataset::SetSaltAndPepperNoiseMaximumDensity ( float saltAndPepperMaximumDensity )

The maximum density of the salt and pepper noise.

The salt and pepper noise sets the value of a number of randomly selected (between EClassificationDataset::SetSaltAndPepperNoiseMinimumDensity and EClassificationDataset::SetSaltAndPepperNoiseMaximumDensity) pixels to its minimum or maximum value.

Its value must be between EClassificationDataset::SetSaltAndPepperNoiseMinimumDensity and 1.

Parameters

saltAndPepperMaximumDensity The maximum density of the salt and pepper noise expressed in percentage of the total number of pixels of the image

This noise is computed after all the other noises.

◆ SetSaltAndPepperNoiseMinimumDensity()

void eVision::EasyDeepLearning::EClassificationDataset::SetSaltAndPepperNoiseMinimumDensity ( float saltAndPepperMinimumDensity )

The minimum density of the salt and pepper noise.

The salt and pepper noise sets the value of a number of randomly selected (between EClassificationDataset::SetSaltAndPepperNoiseMinimumDensity and EClassificationDataset::SetSaltAndPepperNoiseMaximumDensity) pixels to its minimum or maximum value.

Its value must be between 0 and EClassificationDataset::SetSaltAndPepperNoiseMaximumDensity.

Parameters

saltAndPepperMinimumDensity The minimum density of the salt and pepper noise expressed in percentage of the total number of pixels of the image

This noise is computed after all the other noises.

◆ SetSpeckleNoiseMaximumStandardDeviation()

void eVision::EasyDeepLearning::EClassificationDataset::SetSpeckleNoiseMaximumStandardDeviation ( float speckleMaximumDeviation )

The speckle noise maximum standard deviation.

The speckle noise is a multiplicative noise sampled from a Gamma distribution with a mean of 1 and with its standard deviation between EClassificationDataset::SetSpeckleNoiseMinimumStandardDeviation and EClassificationDataset::SetSpeckleNoiseMaximumStandardDeviation.

Its value must be strictly higher than EClassificationDataset::SetSpeckleNoiseMinimumStandardDeviation.

Parameters

speckleMaximumDeviation The standard maximum deviation of the gamma distribution

This noise is computed before the salt and paper noise.

◆ SetSpeckleNoiseMinimumStandardDeviation()

void eVision::EasyDeepLearning::EClassificationDataset::SetSpeckleNoiseMinimumStandardDeviation ( float speckleMinimumDeviation )

The speckle noise minimum standard deviation.

The speckle noise is a multiplicative noise sampled from a Gamma distribution with a mean of 1 and with its standard deviation between EClassificationDataset::SetSpeckleNoiseMinimumStandardDeviation and EClassificationDataset::SetSpeckleNoiseMaximumStandardDeviation.

Its value must be strictly positive and lower than EClassificationDataset::SetSpeckleNoiseMaximumStandardDeviation.

Parameters

speckleMinimumDeviation The standard minimum deviation of the gamma distribution

This noise is computed before the salt and paper noise.

◆ SplitDataset() [1/3]

void eVision::EasyDeepLearning::EClassificationDataset::SplitDataset	(	eVision::EasyDeepLearning::EClassificationDataset &	d1,
		eVision::EasyDeepLearning::EClassificationDataset &	d2 ) const

Splits the dataset in two parts to be used for training and validation respectively.

Parameters

d1	First part of the dataset
d2	Second part of the dataset

◆ SplitDataset() [2/3]

void eVision::EasyDeepLearning::EClassificationDataset::SplitDataset	(	eVision::EasyDeepLearning::EClassificationDataset &	d1,
		eVision::EasyDeepLearning::EClassificationDataset &	d2,
		float	proportion ) const

Splits the dataset in two parts to be used for training and validation respectively.

Parameters

d1	First part of the dataset
d2	Second part of the dataset
proportion	Proportion of image of each class to put into the first part. The remaining images are put in d2

◆ SplitDataset() [3/3]

void eVision::EasyDeepLearning::EClassificationDataset::SplitDataset	(	eVision::EasyDeepLearning::EClassificationDataset &	d1,
		eVision::EasyDeepLearning::EClassificationDataset &	d2,
		float	proportion,
		bool	random ) const

Splits the dataset in two parts to be used for training and validation respectively.

Parameters

d1	First part of the dataset
d2	Second part of the dataset
proportion	Proportion of image of each class to put into the first part. The remaining images are put in d2
random	Randomly sample the images.

◆ SplitDatasetForLocator() [1/3]

void eVision::EasyDeepLearning::EClassificationDataset::SplitDatasetForLocator	(	eVision::EasyDeepLearning::EClassificationDataset &	d1,
		eVision::EasyDeepLearning::EClassificationDataset &	d2 ) const

Splits the dataset in two parts for training and validation of a ELocator tool. Images without object labeling are excluded from the split.

Parameters

d1	First part of the dataset
d2	Second part of the dataset

The method ensures that all the object labels are represented in d1. Thus, even when the parameter random is set to false, the images in d1 and d2 can be ordered differently than they were in the original dataset.

◆ SplitDatasetForLocator() [2/3]

void eVision::EasyDeepLearning::EClassificationDataset::SplitDatasetForLocator	(	eVision::EasyDeepLearning::EClassificationDataset &	d1,
		eVision::EasyDeepLearning::EClassificationDataset &	d2,
		float	proportion ) const

Splits the dataset in two parts for training and validation of a ELocator tool. Images without object labeling are excluded from the split.

Parameters

d1	First part of the dataset
d2	Second part of the dataset
proportion	Proportion of image of each class to put into the first part. The remaining images are put in d2

The method ensures that all the object labels are represented in d1. Thus, even when the parameter random is set to false, the images in d1 and d2 can be ordered differently than they were in the original dataset.

◆ SplitDatasetForLocator() [3/3]

void eVision::EasyDeepLearning::EClassificationDataset::SplitDatasetForLocator	(	eVision::EasyDeepLearning::EClassificationDataset &	d1,
		eVision::EasyDeepLearning::EClassificationDataset &	d2,
		float	proportion,
		bool	random ) const

Splits the dataset in two parts for training and validation of a ELocator tool. Images without object labeling are excluded from the split.

Parameters

d1	First part of the dataset
d2	Second part of the dataset
proportion	Proportion of image of each class to put into the first part. The remaining images are put in d2
random	Randomly sample the images.

The method ensures that all the object labels are represented in d1. Thus, even when the parameter random is set to false, the images in d1 and d2 can be ordered differently than they were in the original dataset.

◆ SplitDatasetForSegmentation() [1/3]

void eVision::EasyDeepLearning::EClassificationDataset::SplitDatasetForSegmentation	(	eVision::EasyDeepLearning::EClassificationDataset &	d1,
		eVision::EasyDeepLearning::EClassificationDataset &	d2 ) const

Splits the dataset in two parts for a supervised segmenter. The two parts are to be used for training and validation respectively.

Parameters

d1	First part of the dataset
d2	Second part of the dataset

The method ensures that all the segmentation labels are represented in d1. Thus, even when the parameter random is set to false, the images in d1 and d2 can be ordered differently than they were in the original dataset.

◆ SplitDatasetForSegmentation() [2/3]

void eVision::EasyDeepLearning::EClassificationDataset::SplitDatasetForSegmentation	(	eVision::EasyDeepLearning::EClassificationDataset &	d1,
		eVision::EasyDeepLearning::EClassificationDataset &	d2,
		float	proportion ) const

Splits the dataset in two parts for a supervised segmenter. The two parts are to be used for training and validation respectively.

Parameters

d1	First part of the dataset
d2	Second part of the dataset
proportion	Proportion of image of each class to put into the first part. The remaining images are put in d2

The method ensures that all the segmentation labels are represented in d1. Thus, even when the parameter random is set to false, the images in d1 and d2 can be ordered differently than they were in the original dataset.

◆ SplitDatasetForSegmentation() [3/3]

void eVision::EasyDeepLearning::EClassificationDataset::SplitDatasetForSegmentation	(	eVision::EasyDeepLearning::EClassificationDataset &	d1,
		eVision::EasyDeepLearning::EClassificationDataset &	d2,
		float	proportion,
		bool	random ) const

Splits the dataset in two parts for a supervised segmenter. The two parts are to be used for training and validation respectively.

Parameters

d1	First part of the dataset
d2	Second part of the dataset
proportion	Proportion of image of each class to put into the first part. The remaining images are put in d2
random	Randomly sample the images.

The method ensures that all the segmentation labels are represented in d1. Thus, even when the parameter random is set to false, the images in d1 and d2 can be ordered differently than they were in the original dataset.

Public Member Functions

Additional Inherited Members

Detailed Description

Constructor & Destructor Documentation

◆ EClassificationDataset() [1/4]

◆ EClassificationDataset() [2/4]

◆ EClassificationDataset() [3/4]

◆ ~EClassificationDataset()

◆ EClassificationDataset() [4/4]

Member Function Documentation

◆ Clear()

◆ ExtractSplit()

◆ GetEnableDataAugmentation()

◆ GetEnableHorizontalFlip()

◆ GetEnableVerticalFlip()

◆ GetGaussianNoiseMaximumStandardDeviation()

◆ GetGaussianNoiseMinimumStandardDeviation()

◆ GetImageCopyWithDataAugmentation() [1/4]

◆ GetImageCopyWithDataAugmentation() [2/4]

◆ GetImageCopyWithDataAugmentation() [3/4]

◆ GetImageCopyWithDataAugmentation() [4/4]

◆ GetImages() [1/2]

◆ GetImages() [2/2]

◆ GetMaxBrightnessOffset()

◆ GetMaxContrastGain()

◆ GetMaxGamma()

◆ GetMaxHorizontalShear()

◆ GetMaxHorizontalShift()

◆ GetMaxHueOffset()

◆ GetMaxRotationAngle()

◆ GetMaxSaturationGain()

◆ GetMaxScale()

◆ GetMaxVerticalShear()

◆ GetMaxVerticalShift()

◆ GetMinContrastGain()

◆ GetMinGamma()

◆ GetMinSaturationGain()

◆ GetMinScale()

◆ GetSaltAndPepperNoiseMaximumDensity()

◆ GetSaltAndPepperNoiseMinimumDensity()

◆ GetSpeckleNoiseMaximumStandardDeviation()

◆ GetSpeckleNoiseMinimumStandardDeviation()

◆ operator=()

◆ SetEnableDataAugmentation()

◆ SetEnableHorizontalFlip()

◆ SetEnableVerticalFlip()

◆ SetGaussianNoiseMaximumStandardDeviation()

◆ SetGaussianNoiseMinimumStandardDeviation()

◆ SetMaxBrightnessOffset()

◆ SetMaxContrastGain()

◆ SetMaxGamma()

◆ SetMaxHorizontalShear()

◆ SetMaxHorizontalShift()

◆ SetMaxHueOffset()

◆ SetMaxRotationAngle()

◆ SetMaxSaturationGain()

◆ SetMaxScale()

◆ SetMaxVerticalShear()

◆ SetMaxVerticalShift()

◆ SetMinContrastGain()

◆ SetMinGamma()

◆ SetMinSaturationGain()

◆ SetMinScale()

◆ SetSaltAndPepperNoiseMaximumDensity()

◆ SetSaltAndPepperNoiseMinimumDensity()

◆ SetSpeckleNoiseMaximumStandardDeviation()

◆ SetSpeckleNoiseMinimumStandardDeviation()

◆ SplitDataset() [1/3]

◆ SplitDataset() [2/3]

◆ SplitDataset() [3/3]

◆ SplitDatasetForLocator() [1/3]

◆ SplitDatasetForLocator() [2/3]

◆ SplitDatasetForLocator() [3/3]

◆ SplitDatasetForSegmentation() [1/3]

◆ SplitDatasetForSegmentation() [2/3]

◆ SplitDatasetForSegmentation() [3/3]