Predictor that may be used for searching objects. More...

#include <cvb/polimago/search_predictor.hpp>

Public Member Functions
	SearchPredictor (const String &fileName)
	Load a saved Polimago search predictor from a file. More...

	SearchPredictor (SearchPredictor &&) noexcept=default
	Move constructor.

SearchPredictor &	operator= (SearchPredictor &&) noexcept=default
	Move assignment operator.

int	MaxNumResults () const noexcept
	Get the maximum number of results that can be extracted in a GridSearch operation. More...

void	SetMaxNumResults (int maxNumResults)
	Set the maximum number of results that can be extracted in a GridSearch operation. Increasing this value will increase the amount of memory allocated by this SearchPredictor object. More...

int	SampleSize () const noexcept
	Get the sample size that has been used in each training set. More...

double	ExtractionRadius () const noexcept
	Get the radius for extracting positive search instances. More...

ValueRange< Angle >	RotationRange () const noexcept
	Get the range of angles that was covered during classifier training. More...

ValueRange< double >	ScaleRange () const noexcept
	Get the range of scale factors that was covered during classifier training. More...

InvarianceType	Invariances () const noexcept
	Get the invariances that have been trained on this classifier. More...

int	NumClassificationSteps () const noexcept
	Get the number of classification steps defined during training. More...

int	FeatureResolutionStep1And2 () const noexcept
	Get the feature resolution to be used for the first two classification steps. More...

int	FeatureResolutionRest () const noexcept
	Get the feature resolution to be used for the third and later classification steps. More...

std::vector< SearchResult >	GridSearch (const Image &img, Rect< int > aoi, double gridStep, double threshold, double locality, int &numCalls)
	Perform a grid search. More...

std::vector< SearchResult >	GridSearch (const Image &img, Rect< int > aoi, double gridStep, double threshold, double locality)
	Perform a grid search. More...

bool	Inspect (const Image &img, SearchResult &res, int &searchDepth)
	Carries out the operation that GridSearch executes for a grid point, starting at the perspective and position defined by the initial value of the parameter SearchResult. More...

bool	Inspect (const Image &img, SearchResult &res, int &searchDepth, std::vector< SearchResult > &trace)
	Carries out the operation that GridSearch executes for a grid point, starting at the perspective and position defined by the initial value of the parameter SearchResult. More...

std::unique_ptr< Image >	SearchResultToImage (const Image &sourceImage, SearchResult res)
	Create a visual representation of a search result. More...

Public Member Functions inherited from PredictorBase
bool	IsCompatible (const Image &img, Point2D< int > pos) const
	Verify the compatibility of a CVB image with this classifier. More...

bool	IsCompatible (const Image &img) const
	Verify the compatibility of a CVB image with this classifier. More...

Public Member Functions inherited from PolimagoFactoryCreatedObject
void *	Handle () const noexcept
	Classic API Polimago handle. More...

String	FileName () const
	Name of the file the object has been loaded from (or empty string if the object was not loaded). More...

void	Save (const String &fileName) const
	Save this object into a file. More...

double	Lambda () const noexcept
	Regularization value that has been used for generating this object. More...

double	Offset () const noexcept
	Intercept weight that has been used for generating this object. More...

InterpolationType	Interpolation () const noexcept
	Interpolation setting used for generating this object. More...

int	ImagePlanes () const noexcept
	The plane count of the images that have been used for generating this classifier. Image on which this classifier is to be used will need to have the same plane count. More...

Rect< int >	FeatureWindowExtent () const noexcept
	The feature window extent that has been used during classifier training. More...

Size2D< int >	RetinaSize () const noexcept
	Size of the 'Retina' in pixels. The retina is the set of paxels onto which the input image is projected using the preprocessing code prior to application/training of the classifier. More...

Point2D< double >	Correction () const noexcept
	Correction factors in X and Y direction required to rescale the input images for projection onto the retina. More...

String	Preprocessing () const
	Preprocessing code with which this object was generated. More...

Static Public Member Functions
static std::unique_ptr< SearchPredictor >	FromHandle (ReleaseObjectGuard &&guard)
	Creates predictor from a classic API handle. More...

static std::unique_ptr< SearchPredictor >	Load (const String &fileName)
	Load a saved predictor from a file. More...

Detailed Description

Predictor that may be used for searching objects.

Constructor & Destructor Documentation

◆ SearchPredictor()

SearchPredictor ( const String & fileName )

inline

Load a saved Polimago search predictor from a file.

Parameters

[in] fileName Name of the file to be loaded.

Returns: Loaded predictor.

Exceptions

Any	exception derived from std::exception including CvbException.

Member Function Documentation

◆ ExtractionRadius()

double ExtractionRadius ( ) const

inlinenoexcept

Get the radius for extracting positive search instances.

Returns: Extraction radius.

Exceptions

Does	not throw any exception.

The unit size is the size of the feature window, i.e. a value of 0.5 means that positive samples can be extracted from withing a range of +/- Feature Window Width and +/- Feature Window Height around each trained instance. Smaller values will necessitate a finer search grid but may yield better results on difficult search tasks.

◆ FeatureResolutionRest()

int FeatureResolutionRest ( ) const

inlinenoexcept

Get the feature resolution to be used for the third and later classification steps.

Returns: Feature resolution.

Exceptions

Does	not throw any exception.

◆ FeatureResolutionStep1And2()

int FeatureResolutionStep1And2 ( ) const

inlinenoexcept

Get the feature resolution to be used for the first two classification steps.

Returns: Feature resolution.

Exceptions

Does	not throw any exception.

◆ FromHandle()

static std::unique_ptr< SearchPredictor > FromHandle ( ReleaseObjectGuard && guard )

inlinestatic

Creates predictor from a classic API handle.

Parameters

[in] guard Life time guard for C-API handle.

Returns: The predictor created from the classic API handle.

Exceptions

Any	exception derived from std::exception including CvbException.

The predictor takes ownership of the handle, so you must share it before using this function.

◆ GridSearch() [1/2]

std::vector< SearchResult > GridSearch	(	const Image &	img,
		Rect< int >	aoi,
		double	gridStep,
		double	threshold,
		double	locality
	)

inline

Perform a grid search.

Parameters

[in]	img	Image to search in.
[in]	aoi	AOI inside the image to search in.
[in]	gridStep	Spacing of the grid points in units of the feature window width/height.
[in]	threshold	Minimum quality a result must have for being reported.
[in]	locality	Minimum distance between the results that this SearchPredictor found.

Returns: List of results that this SearchPredictor found in the area of interest.

Exceptions

Any	exception derived from std::exception including CvbException.

If two results that exceed the quality requirement defined by the threshold parameter are spaced less than 'locality' pixels apart (in either x or y direction) then the result with the lower quality will be eliminated from the result list.

◆ GridSearch() [2/2]

std::vector< SearchResult > GridSearch	(	const Image &	img,
		Rect< int >	aoi,
		double	gridStep,
		double	threshold,
		double	locality,
		int &	numCalls
	)

inline

Perform a grid search.

Parameters

[in]	img	Image to search in.
[in]	aoi	AOI inside the image to search in.
[in]	gridStep	Spacing of the grid points in units of the feature window width/height.
[in]	threshold	Minimum quality a result must have for being reported.
[in]	locality	Minimum distance between the results that this SearchPredictor found.
[out]	numCalls	Will receive the number of calls to the predictor's classification routine that have been carried out in the GridSearch call (output).

Returns: List of results that this SearchPredictor found in the area of interest.

Exceptions

Any	exception derived from std::exception including CvbException.

If two results that exceed the quality requirement defined by the threshold parameter are spaced less than 'locality' pixels apart (in either x or y direction) then the result with the lower quality will be eliminated from the result list.

◆ Inspect() [1/2]

bool Inspect	(	const Image &	img,
		SearchResult &	res,
		int &	searchDepth
	)

inline

Carries out the operation that GridSearch executes for a grid point, starting at the perspective and position defined by the initial value of the parameter SearchResult.

Parameters

[in]	img	Image to work on.
[in,out]	res	Search Result to start with (upon exit: final result).
[in,out]	searchDepth	Maximum number of steps to calculate (upon exit: number of steps actually calculated).

Returns: True if the SearchPredictor found a result when starting at the position defined through the res parameter, false otherwise.

Exceptions

Any	exception derived from std::exception including CvbException.

When the function returns successfully, SearchResult will contain the final perspective and result quality. For additional information please see the tutorial on search functions of Polimago.

◆ Inspect() [2/2]

bool Inspect	(	const Image &	img,
		SearchResult &	res,
		int &	searchDepth,
		std::vector< SearchResult > &	trace
	)

inline

Carries out the operation that GridSearch executes for a grid point, starting at the perspective and position defined by the initial value of the parameter SearchResult.

Parameters

[in]	img	Image to work on.
[in,out]	res	Search Result to start with (upon exit: final result).
[in,out]	searchDepth	Maximum number of steps to calculate (upon exit: number of steps actually calculated).
[out]	trace	Search results that lead from the initial value of the res parameter to the final value.

Returns: True if the SearchPredictor found a result when starting at the position defined through the res parameter, false otherwise.

Exceptions

Any	exception derived from std::exception including CvbException.

When the function returns successfully, SearchResult will contain the final perspective and result quality. For additional information please see the tutorial on search functions of Polimago.

◆ Invariances()

InvarianceType Invariances ( ) const

inlinenoexcept

Get the invariances that have been trained on this classifier.

Returns: Invariance type.

Exceptions

Does	not throw any exception.

◆ Load()

static std::unique_ptr< SearchPredictor > Load ( const String & fileName )

inlinestatic

Load a saved predictor from a file.

Parameters

[in] fileName Name of the file to be loaded.

Returns: Pointer to the loaded predictor.

Exceptions

Any	exception derived from std::exception including CvbException.

Examples: Polimago/QmlCookieClassification.

◆ MaxNumResults()

int MaxNumResults ( ) const

inlinenoexcept

Get the maximum number of results that can be extracted in a GridSearch operation.

Returns: Maximum number of results.

Exceptions

Does	not throw any exception.

◆ NumClassificationSteps()

int NumClassificationSteps ( ) const

inlinenoexcept

Get the number of classification steps defined during training.

Returns: Number of classification steps.

Exceptions

Does	not throw any exception.

◆ RotationRange()

ValueRange< Angle > RotationRange ( ) const

inlinenoexcept

Get the range of angles that was covered during classifier training.

Returns: Rotation range.

Exceptions

Does	not throw any exception.

◆ SampleSize()

int SampleSize ( ) const

inlinenoexcept

Get the sample size that has been used in each training set.

Returns: Sample size.

Exceptions

Does	not throw any exception.

◆ ScaleRange()

ValueRange< double > ScaleRange ( ) const

inlinenoexcept

Get the range of scale factors that was covered during classifier training.

Returns: Scale range.

Exceptions

Does	not throw any exception.

◆ SearchResultToImage()

std::unique_ptr< Image > SearchResultToImage	(	const Image &	sourceImage,
		SearchResult	res
	)

inline

Create a visual representation of a search result.

Parameters

[in]	sourceImage	Image to create the representation from. This should be the image on which the result was found with this classifier.
[in]	res	Search result (should match the image and this classifier).

Returns: Visual representation of the search result.

Exceptions

Any	exception derived from std::exception including CvbException.

◆ SetMaxNumResults()

void SetMaxNumResults ( int maxNumResults )

inline

Set the maximum number of results that can be extracted in a GridSearch operation. Increasing this value will increase the amount of memory allocated by this SearchPredictor object.

Parameters

[in] maxNumResults The maximum number of results.

Exceptions

Any	exception derived from std::exception including CvbException.

Public Member Functions

Static Public Member Functions

Detailed Description

Constructor & Destructor Documentation

◆ SearchPredictor()

Member Function Documentation

◆ ExtractionRadius()

◆ FeatureResolutionRest()

◆ FeatureResolutionStep1And2()

◆ FromHandle()

◆ GridSearch() [1/2]

◆ GridSearch() [2/2]

◆ Inspect() [1/2]

◆ Inspect() [2/2]

◆ Invariances()

◆ Load()

◆ MaxNumResults()

◆ NumClassificationSteps()

◆ RotationRange()

◆ SampleSize()

◆ ScaleRange()

◆ SearchResultToImage()

◆ SetMaxNumResults()