cv::LightGlueMatcher Class Reference

LightGlue feature matcher. More...

#include <opencv2/features.hpp>

Public Member Functions
virtual	~LightGlueMatcher ()
virtual void	clearPairInfo ()=0
	Clears stored pair context information.
virtual void	setPairInfo (InputArray queryKpts, InputArray trainKpts, Size queryImageSize=Size(), Size trainImageSize=Size())=0
	Sets the keypoint and image size context for the next match() call.
Public Member Functions inherited from cv::DescriptorMatcher
virtual	~DescriptorMatcher ()
virtual void	add (InputArrayOfArrays descriptors)
	Adds descriptors to train a CPU(trainDescCollectionis) or GPU(utrainDescCollectionis) descriptor collection.
virtual void	clear () CV_OVERRIDE
	Clears the train descriptor collections.
virtual CV_NODISCARD_STD Ptr< DescriptorMatcher >	clone (bool emptyTrainData=false) const =0
	Clones the matcher.
virtual bool	empty () const CV_OVERRIDE
	Returns true if there are no train descriptors in the both collections.
const std::vector< Mat > &	getTrainDescriptors () const
	Returns a constant link to the train descriptor collection trainDescCollection .
virtual bool	isMaskSupported () const =0
	Returns true if the descriptor matcher supports masking permissible matches.
void	knnMatch (InputArray queryDescriptors, InputArray trainDescriptors, std::vector< std::vector< DMatch > > &matches, int k, InputArray mask=noArray(), bool compactResult=false) const
	Finds the k best matches for each descriptor from a query set.
void	knnMatch (InputArray queryDescriptors, std::vector< std::vector< DMatch > > &matches, int k, InputArrayOfArrays masks=noArray(), bool compactResult=false)
void	match (InputArray queryDescriptors, InputArray trainDescriptors, std::vector< DMatch > &matches, InputArray mask=noArray()) const
	Finds the best match for each descriptor from a query set.
void	match (InputArray queryDescriptors, std::vector< DMatch > &matches, InputArrayOfArrays masks=noArray())
void	radiusMatch (InputArray queryDescriptors, InputArray trainDescriptors, std::vector< std::vector< DMatch > > &matches, float maxDistance, InputArray mask=noArray(), bool compactResult=false) const
	For each query descriptor, finds the training descriptors not farther than the specified distance.
void	radiusMatch (InputArray queryDescriptors, std::vector< std::vector< DMatch > > &matches, float maxDistance, InputArrayOfArrays masks=noArray(), bool compactResult=false)
virtual void	read (const FileNode &) CV_OVERRIDE
	Reads algorithm parameters from a file storage.
void	read (const String &fileName)
virtual void	train ()
	Trains a descriptor matcher.
void	write (const Ptr< FileStorage > &fs, const String &name) const
void	write (const String &fileName) const
virtual void	write (FileStorage &) const CV_OVERRIDE
	Stores algorithm parameters in a file storage.
void	write (FileStorage &fs, const String &name) const
Public Member Functions inherited from cv::Algorithm
	Algorithm ()
virtual	~Algorithm ()
virtual String	getDefaultName () const
virtual void	save (const String &filename) const
void	write (FileStorage &fs, const String &name) const

Static Public Member Functions
static Ptr< LightGlueMatcher >	create (const String &modelPath, float scoreThreshold=0.0f, int backend=0, int target=0)
	Creates LightGlueMatcher from a model file path.
Static Public Member Functions inherited from cv::DescriptorMatcher
static Ptr< DescriptorMatcher >	create (const DescriptorMatcher::MatcherType &matcherType)
static Ptr< DescriptorMatcher >	create (const String &descriptorMatcherType)
	Creates a descriptor matcher of a given type with the default parameters (using default constructor).
Static Public Member Functions inherited from cv::Algorithm
template<typename _Tp>
static Ptr< _Tp >	load (const String &filename, const String &objname=String())
	Loads algorithm from the file.
template<typename _Tp>
static Ptr< _Tp >	loadFromString (const String &strModel, const String &objname=String())
	Loads algorithm from a String.
template<typename _Tp>
static Ptr< _Tp >	read (const FileNode &fn)
	Reads algorithm from the file node.

Protected Member Functions
	LightGlueMatcher ()
Protected Member Functions inherited from cv::DescriptorMatcher
void	checkMasks (InputArrayOfArrays masks, int queryDescriptorsCount) const
virtual void	knnMatchImpl (InputArray queryDescriptors, std::vector< std::vector< DMatch > > &matches, int k, InputArrayOfArrays masks=noArray(), bool compactResult=false)=0
virtual void	radiusMatchImpl (InputArray queryDescriptors, std::vector< std::vector< DMatch > > &matches, float maxDistance, InputArrayOfArrays masks=noArray(), bool compactResult=false)=0
Protected Member Functions inherited from cv::Algorithm
void	writeFormat (FileStorage &fs) const

Additional Inherited Members
Public Types inherited from cv::DescriptorMatcher
enum	MatcherType {   FLANNBASED = 1 ,   BRUTEFORCE = 2 ,   BRUTEFORCE_L1 = 3 ,   BRUTEFORCE_HAMMING = 4 ,   BRUTEFORCE_HAMMINGLUT = 5 ,   BRUTEFORCE_SL2 = 6 }
Static Protected Member Functions inherited from cv::DescriptorMatcher
static CV_NODISCARD_STD Mat	clone_op (Mat m)
static bool	isMaskedOut (InputArrayOfArrays masks, int queryIdx)
static bool	isPossibleMatch (InputArray mask, int queryIdx, int trainIdx)
Protected Attributes inherited from cv::DescriptorMatcher
std::vector< Mat >	trainDescCollection
	Collection of descriptors from train images.
std::vector< UMat >	utrainDescCollection

Detailed Description

LightGlue feature matcher.

LightGlue is a CNN-based feature matcher, as described in [Lindenberger23] . It takes keypoint locations and descriptors from two images and directly predicts match pairs. Unlike traditional matchers that compute descriptor distances, LightGlue uses attention mechanisms to produce confidence scores for each potential match pair.

The matcher extends DescriptorMatcher and supports the standard match(), knnMatch(), and radiusMatch() interfaces. Context (keypoints and image sizes) must be provided via setPairInfo() before matching.

Constructor & Destructor Documentation

LightGlueMatcher()

cv::LightGlueMatcher::LightGlueMatcher ( )

protected

~LightGlueMatcher()

virtual cv::LightGlueMatcher::~LightGlueMatcher ( )

virtual

Member Function Documentation

clearPairInfo()

virtual void cv::LightGlueMatcher::clearPairInfo ( )

pure virtual

Python:
	cv.LightGlueMatcher.clearPairInfo(		) ->	None

Clears stored pair context information.

create()

Ptr< LightGlueMatcher > cv::LightGlueMatcher::create ( const String & modelPath, float scoreThreshold = 0.0f, int backend = 0, int target = 0 )

static

Python:
	cv.LightGlueMatcher.create(	modelPath[, scoreThreshold[, backend[, target]]]	) ->	retval
	cv.LightGlueMatcher.createFromMemory(	modelData[, scoreThreshold[, backend[, target]]]	) ->	retval
	cv.LightGlueMatcher_create(	modelPath[, scoreThreshold[, backend[, target]]]	) ->	retval
	cv.LightGlueMatcher_createFromMemory(	modelData[, scoreThreshold[, backend[, target]]]	) ->	retval

Creates LightGlueMatcher from a model file path.

Parameters

modelPath	Path to the ONNX model file.
scoreThreshold	Match confidence threshold.
backend	DNN backend
target	DNN target

Examples

samples/cpp/stitching_detailed.cpp.

setPairInfo()

virtual void cv::LightGlueMatcher::setPairInfo ( InputArray queryKpts, InputArray trainKpts, Size queryImageSize = Size(), Size trainImageSize = Size() )

pure virtual

Python:
	cv.LightGlueMatcher.setPairInfo(	queryKpts, trainKpts[, queryImageSize[, trainImageSize]]	) ->	None

Sets the keypoint and image size context for the next match() call.

This provides the spatial context that LightGlue needs in addition to descriptors. Must be called before match()/knnMatch()/radiusMatch() unless using automatic context from in-process ALIKED instances.

Parameters

queryKpts	Query image keypoints (Nx2 float matrix with x,y coordinates).
trainKpts	Train image keypoints (Nx2 float matrix with x,y coordinates).
queryImageSize	Size of the query image (width, height).
trainImageSize	Size of the train image (width, height).

---

The documentation for this class was generated from the following file:

• opencv2/features.hpp