Class implementing PCT (position-color-texture) signature extraction as described in [KrulisLS16]. The algorithm is divided to a feature sampler and a clusterizer. Feature sampler produces samples at given set of coordinates. Clusterizer then produces clusters of these samples using k-means algorithm. Resulting set of clusters is the signature of the input image. [詳解]

#include <xfeatures2d.hpp>

cv::Algorithmを継承しています。

公開型
enum	DistanceFunction { L0_25 , L0_5 , L1 , L2 , L2SQUARED , L5 , L_INFINITY }
	Lp distance function selector.

enum	PointDistribution { UNIFORM , REGULAR , NORMAL }
	Point distributions supported by random point generator. [詳解]

enum	SimilarityFunction { MINUS , GAUSSIAN , HEURISTIC }
	Similarity function selector. [詳解]

公開メンバ関数
virtual CV_WRAP void	computeSignature (InputArray image, OutputArray signature) const =0
	Computes signature of given image. [詳解]

virtual CV_WRAP void	computeSignatures (const std::vector< Mat > &images, std::vector< Mat > &signatures) const =0
	Computes signatures for multiple images in parallel. [詳解]

virtual CV_WRAP int	getSampleCount () const =0
	Number of initial samples taken from the image.

virtual CV_WRAP int	getGrayscaleBits () const =0
	Color resolution of the greyscale bitmap represented in allocated bits (i.e., value 4 means that 16 shades of grey are used). The greyscale bitmap is used for computing contrast and entropy values.

virtual CV_WRAP void	setGrayscaleBits (int grayscaleBits)=0
	Color resolution of the greyscale bitmap represented in allocated bits (i.e., value 4 means that 16 shades of grey are used). The greyscale bitmap is used for computing contrast and entropy values.

virtual CV_WRAP int	getWindowRadius () const =0
	Size of the texture sampling window used to compute contrast and entropy (center of the window is always in the pixel selected by x,y coordinates of the corresponding feature sample).

virtual CV_WRAP void	setWindowRadius (int radius)=0
	Size of the texture sampling window used to compute contrast and entropy (center of the window is always in the pixel selected by x,y coordinates of the corresponding feature sample).

virtual CV_WRAP float	getWeightX () const =0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP void	setWeightX (float weight)=0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP float	getWeightY () const =0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP void	setWeightY (float weight)=0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP float	getWeightL () const =0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP void	setWeightL (float weight)=0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP float	getWeightA () const =0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP void	setWeightA (float weight)=0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP float	getWeightB () const =0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP void	setWeightB (float weight)=0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP float	getWeightContrast () const =0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP void	setWeightContrast (float weight)=0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP float	getWeightEntropy () const =0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP void	setWeightEntropy (float weight)=0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space (x,y = position; L,a,b = color in CIE Lab space; c = contrast. e = entropy)

virtual CV_WRAP std::vector< Point2f >	getSamplingPoints () const =0
	Initial samples taken from the image. These sampled features become the input for clustering.

virtual CV_WRAP void	setWeight (int idx, float value)=0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space. [詳解]

virtual CV_WRAP void	setWeights (const std::vector< float > &weights)=0
	Weights (multiplicative constants) that linearly stretch individual axes of the feature space. [詳解]

virtual CV_WRAP void	setTranslation (int idx, float value)=0
	Translations of the individual axes of the feature space. [詳解]

virtual CV_WRAP void	setTranslations (const std::vector< float > &translations)=0
	Translations of the individual axes of the feature space. [詳解]

virtual CV_WRAP void	setSamplingPoints (std::vector< Point2f > samplingPoints)=0
	Sets sampling points used to sample the input image. [詳解]

virtual CV_WRAP std::vector< int >	getInitSeedIndexes () const =0
	Initial seeds (initial number of clusters) for the k-means algorithm.

virtual CV_WRAP void	setInitSeedIndexes (std::vector< int > initSeedIndexes)=0
	Initial seed indexes for the k-means algorithm.

virtual CV_WRAP int	getInitSeedCount () const =0
	Number of initial seeds (initial number of clusters) for the k-means algorithm.

virtual CV_WRAP int	getIterationCount () const =0
	Number of iterations of the k-means clustering. We use fixed number of iterations, since the modified clustering is pruning clusters (not iteratively refining k clusters).

virtual CV_WRAP void	setIterationCount (int iterationCount)=0
	Number of iterations of the k-means clustering. We use fixed number of iterations, since the modified clustering is pruning clusters (not iteratively refining k clusters).

virtual CV_WRAP int	getMaxClustersCount () const =0
	Maximal number of generated clusters. If the number is exceeded, the clusters are sorted by their weights and the smallest clusters are cropped.

virtual CV_WRAP void	setMaxClustersCount (int maxClustersCount)=0
	Maximal number of generated clusters. If the number is exceeded, the clusters are sorted by their weights and the smallest clusters are cropped.

virtual CV_WRAP int	getClusterMinSize () const =0
	This parameter multiplied by the index of iteration gives lower limit for cluster size. Clusters containing fewer points than specified by the limit have their centroid dismissed and points are reassigned.

virtual CV_WRAP void	setClusterMinSize (int clusterMinSize)=0
	This parameter multiplied by the index of iteration gives lower limit for cluster size. Clusters containing fewer points than specified by the limit have their centroid dismissed and points are reassigned.

virtual CV_WRAP float	getJoiningDistance () const =0
	Threshold euclidean distance between two centroids. If two cluster centers are closer than this distance, one of the centroid is dismissed and points are reassigned.

virtual CV_WRAP void	setJoiningDistance (float joiningDistance)=0
	Threshold euclidean distance between two centroids. If two cluster centers are closer than this distance, one of the centroid is dismissed and points are reassigned.

virtual CV_WRAP float	getDropThreshold () const =0
	Remove centroids in k-means whose weight is lesser or equal to given threshold.

virtual CV_WRAP void	setDropThreshold (float dropThreshold)=0
	Remove centroids in k-means whose weight is lesser or equal to given threshold.

virtual CV_WRAP int	getDistanceFunction () const =0
	Distance function selector used for measuring distance between two points in k-means.

virtual CV_WRAP void	setDistanceFunction (int distanceFunction)=0
	Distance function selector used for measuring distance between two points in k-means. Available: L0_25, L0_5, L1, L2, L2SQUARED, L5, L_INFINITY.

基底クラス cv::Algorithm に属する継承公開メンバ関数
virtual CV_WRAP void	clear ()
	Clears the algorithm state [詳解]

virtual void	write (FileStorage &fs) const
	Stores algorithm parameters in a file storage [詳解]

CV_WRAP void	write (const Ptr< FileStorage > &fs, const String &name=String()) const
	simplified API for language bindings これはオーバーロードされたメンバ関数です。利便性のために用意されています。元の関数との違いは引き数のみです。

virtual CV_WRAP void	read (const FileNode &fn)
	Reads algorithm parameters from a file storage [詳解]

virtual CV_WRAP bool	empty () const
	Returns true if the Algorithm is empty (e.g. in the very beginning or after unsuccessful read [詳解]

virtual CV_WRAP void	save (const String &filename) const

virtual CV_WRAP String	getDefaultName () const

静的公開メンバ関数
static CV_WRAP Ptr< PCTSignatures >	create (const int initSampleCount=2000, const int initSeedCount=400, const int pointDistribution=0)
	Creates PCTSignatures algorithm using sample and seed count. It generates its own sets of sampling points and clusterization seed indexes. [詳解]

static CV_WRAP Ptr< PCTSignatures >	create (const std::vector< Point2f > &initSamplingPoints, const int initSeedCount)
	Creates PCTSignatures algorithm using pre-generated sampling points and number of clusterization seeds. It uses the provided sampling points and generates its own clusterization seed indexes. [詳解]

static CV_WRAP Ptr< PCTSignatures >	create (const std::vector< Point2f > &initSamplingPoints, const std::vector< int > &initClusterSeedIndexes)
	Creates PCTSignatures algorithm using pre-generated sampling points and clusterization seeds indexes. [詳解]

static CV_WRAP void	drawSignature (InputArray source, InputArray signature, OutputArray result, float radiusToShorterSideRatio=1.0/8, int borderThickness=1)
	Draws signature in the source image and outputs the result. Signatures are visualized as a circle with radius based on signature weight and color based on signature color. Contrast and entropy are not visualized. [詳解]

static CV_WRAP void	generateInitPoints (std::vector< Point2f > &initPoints, const int count, int pointDistribution)
	Generates initial sampling points according to selected point distribution. [詳解]

基底クラス cv::Algorithm に属する継承静的公開メンバ関数
template<typename _Tp >
static Ptr< _Tp >	read (const FileNode &fn)
	Reads algorithm from the file node [詳解]

template<typename _Tp >
static Ptr< _Tp >	load (const String &filename, const String &objname=String())
	Loads algorithm from the file [詳解]

template<typename _Tp >
static Ptr< _Tp >	loadFromString (const String &strModel, const String &objname=String())
	Loads algorithm from a String [詳解]

その他の継承メンバ
基底クラス cv::Algorithm に属する継承限定公開メンバ関数
void	writeFormat (FileStorage &fs) const

詳解

Class implementing PCT (position-color-texture) signature extraction as described in [KrulisLS16]. The algorithm is divided to a feature sampler and a clusterizer. Feature sampler produces samples at given set of coordinates. Clusterizer then produces clusters of these samples using k-means algorithm. Resulting set of clusters is the signature of the input image.

A signature is an array of SIGNATURE_DIMENSION-dimensional points. Used dimensions are: weight, x, y position; lab color, contrast, entropy. [KrulisLS16] [BeecksUS10]

列挙型メンバ詳解

◆ PointDistribution

enum cv::xfeatures2d::PCTSignatures::PointDistribution

Point distributions supported by random point generator.

列挙値
UNIFORM	Generate numbers uniformly.
REGULAR	Generate points in a regular grid.
NORMAL	Generate points with normal (gaussian) distribution.

◆ SimilarityFunction

enum cv::xfeatures2d::PCTSignatures::SimilarityFunction

Similarity function selector.

参照: Christian Beecks, Merih Seran Uysal, Thomas Seidl. Signature quadratic form distance. In Proceedings of the ACM International Conference on Image and Video Retrieval, pages 438-445. ACM, 2010. [BeecksUS10]

覚え書き

For selected distance function:

$d(c_i, c_j)$

and parameter:

$\alpha$

列挙値
MINUS	$-d(c_i, c_j)$
GAUSSIAN	$e^{ -\alpha * d^2(c_i, c_j)}$
HEURISTIC	$\frac{1}{\alpha + d(c_i, c_j)}$

関数詳解

◆ computeSignature()

virtual CV_WRAP void cv::xfeatures2d::PCTSignatures::computeSignature	(	InputArray	image,
		OutputArray	signature
	)		const

pure virtual

Computes signature of given image.

引数

image	Input image of CV_8U type.
signature	Output computed signature.

◆ computeSignatures()

virtual CV_WRAP void cv::xfeatures2d::PCTSignatures::computeSignatures	(	const std::vector< Mat > &	images,
		std::vector< Mat > &	signatures
	)		const

pure virtual

Computes signatures for multiple images in parallel.

引数

images	Vector of input images of CV_8U type.
signatures	Vector of computed signatures.

◆ create() [1/3]

static CV_WRAP Ptr< PCTSignatures > cv::xfeatures2d::PCTSignatures::create	(	const int	initSampleCount = `2000`,
		const int	initSeedCount = `400`,
		const int	pointDistribution = `0`
	)

static

Creates PCTSignatures algorithm using sample and seed count. It generates its own sets of sampling points and clusterization seed indexes.

引数

initSampleCount	Number of points used for image sampling.
initSeedCount	Number of initial clusterization seeds. Must be lower or equal to initSampleCount
pointDistribution	Distribution of generated points. Default: UNIFORM. Available: UNIFORM, REGULAR, NORMAL.

戻り値: Created algorithm.

◆ create() [2/3]

static CV_WRAP Ptr< PCTSignatures > cv::xfeatures2d::PCTSignatures::create	(	const std::vector< Point2f > &	initSamplingPoints,
		const int	initSeedCount
	)

static

Creates PCTSignatures algorithm using pre-generated sampling points and number of clusterization seeds. It uses the provided sampling points and generates its own clusterization seed indexes.

引数

initSamplingPoints	Sampling points used in image sampling.
initSeedCount	Number of initial clusterization seeds. Must be lower or equal to initSamplingPoints.size().

戻り値: Created algorithm.

◆ create() [3/3]

static CV_WRAP Ptr< PCTSignatures > cv::xfeatures2d::PCTSignatures::create	(	const std::vector< Point2f > &	initSamplingPoints,
		const std::vector< int > &	initClusterSeedIndexes
	)

static

Creates PCTSignatures algorithm using pre-generated sampling points and clusterization seeds indexes.

引数

initSamplingPoints	Sampling points used in image sampling.
initClusterSeedIndexes	Indexes of initial clusterization seeds. Its size must be lower or equal to initSamplingPoints.size().

戻り値: Created algorithm.

◆ drawSignature()

static CV_WRAP void cv::xfeatures2d::PCTSignatures::drawSignature	(	InputArray	source,
		InputArray	signature,
		OutputArray	result,
		float	radiusToShorterSideRatio = `1.0/8`,
		int	borderThickness = `1`
	)

static

Draws signature in the source image and outputs the result. Signatures are visualized as a circle with radius based on signature weight and color based on signature color. Contrast and entropy are not visualized.

引数

source	Source image.
signature	Image signature.
result	Output result.
radiusToShorterSideRatio	Determines maximal radius of signature in the output image.
borderThickness	Border thickness of the visualized signature.

◆ generateInitPoints()

static CV_WRAP void cv::xfeatures2d::PCTSignatures::generateInitPoints	(	std::vector< Point2f > &	initPoints,
		const int	count,
		int	pointDistribution
	)

static

Generates initial sampling points according to selected point distribution.

引数

initPoints	Output vector where the generated points will be saved.
count	Number of points to generate.
pointDistribution	Point distribution selector. Available: UNIFORM, REGULAR, NORMAL.

覚え書き: Generated coordinates are in range [0..1)

◆ setSamplingPoints()

virtual CV_WRAP void cv::xfeatures2d::PCTSignatures::setSamplingPoints ( std::vector< Point2f > samplingPoints )

pure virtual

Sets sampling points used to sample the input image.

引数

samplingPoints Vector of sampling points in range [0..1)

覚え書き: Number of sampling points must be greater or equal to clusterization seed count.

◆ setTranslation()

virtual CV_WRAP void cv::xfeatures2d::PCTSignatures::setTranslation	(	int	idx,
		float	value
	)

pure virtual

Translations of the individual axes of the feature space.

引数

idx	ID of the translation
value	Value of the translation

覚え書き: WEIGHT_IDX = 0; X_IDX = 1; Y_IDX = 2; L_IDX = 3; A_IDX = 4; B_IDX = 5; CONTRAST_IDX = 6; ENTROPY_IDX = 7;

◆ setTranslations()

virtual CV_WRAP void cv::xfeatures2d::PCTSignatures::setTranslations ( const std::vector< float > & translations )

pure virtual

Translations of the individual axes of the feature space.

引数

translations Values of all translations.

覚え書き: WEIGHT_IDX = 0; X_IDX = 1; Y_IDX = 2; L_IDX = 3; A_IDX = 4; B_IDX = 5; CONTRAST_IDX = 6; ENTROPY_IDX = 7;

◆ setWeight()

virtual CV_WRAP void cv::xfeatures2d::PCTSignatures::setWeight	(	int	idx,
		float	value
	)

pure virtual

Weights (multiplicative constants) that linearly stretch individual axes of the feature space.

引数

idx	ID of the weight
value	Value of the weight

覚え書き: WEIGHT_IDX = 0; X_IDX = 1; Y_IDX = 2; L_IDX = 3; A_IDX = 4; B_IDX = 5; CONTRAST_IDX = 6; ENTROPY_IDX = 7;

◆ setWeights()

virtual CV_WRAP void cv::xfeatures2d::PCTSignatures::setWeights ( const std::vector< float > & weights )

pure virtual

Weights (multiplicative constants) that linearly stretch individual axes of the feature space.

引数

weights Values of all weights.

覚え書き: WEIGHT_IDX = 0; X_IDX = 1; Y_IDX = 2; L_IDX = 3; A_IDX = 4; B_IDX = 5; CONTRAST_IDX = 6; ENTROPY_IDX = 7;

このクラス詳解は次のファイルから抽出されました:

xfeatures2d.hpp

公開型

公開メンバ関数

静的公開メンバ関数

その他の継承メンバ

詳解

列挙型メンバ詳解

◆ PointDistribution

◆ SimilarityFunction

関数詳解

◆ computeSignature()

◆ computeSignatures()

◆ create() [1/3]

◆ create() [2/3]

◆ create() [3/3]

◆ drawSignature()

◆ generateInitPoints()

◆ setSamplingPoints()

◆ setTranslation()

◆ setTranslations()

◆ setWeight()

◆ setWeights()