Class cv::detail::LightGlueFeaturesMatcher#

Features matcher that adapts LightGlueMatcher (DescriptorMatcher) to the stitching pipeline's FeaturesMatcher interface. View details

#include <opencv2/stitching/detail/matchers.hpp>

Collaboration diagram for cv::detail::LightGlueFeaturesMatcher:

cv::detail::LightGlueFeaturesMatcher Node1 cv::detail::LightGlueFeatures Matcher   + LightGlueFeaturesMatcher() + setScoreThreshold() # match() Node2 cv::detail::FeaturesMatcher   + ~FeaturesMatcher() + collectGarbage() + isThreadSafe() + operator()() + operator()() # FeaturesMatcher() # match() # match() Node2->Node1 Node3 bool     Node3->Node2 #is_thread_safe_ Node4 Ptr< cv::LightGlueMatcher >     Node4->Node1 #lgMatcher_ Node5 int     Node5->Node1 #num_matches_thresh1 _ #num_matches_thresh2_ Node6 double     Node6->Node1 #matches_confidence _thresh_ Node7 float     Node7->Node1 #lg_score_thresh_

cv::detail::LightGlueFeaturesMatcher Node1 cv::detail::LightGlueFeatures Matcher   + LightGlueFeaturesMatcher() + setScoreThreshold() # match() Node2 cv::detail::FeaturesMatcher   + ~FeaturesMatcher() + collectGarbage() + isThreadSafe() + operator()() + operator()() # FeaturesMatcher() # match() # match() Node2->Node1 Node3 bool     Node3->Node2 #is_thread_safe_ Node4 Ptr< cv::LightGlueMatcher >     Node4->Node1 #lgMatcher_ Node5 int     Node5->Node1 #num_matches_thresh1 _ #num_matches_thresh2_ Node6 double     Node6->Node1 #matches_confidence _thresh_ Node7 float     Node7->Node1 #lg_score_thresh_

Public Member Functions#

Public Member Functions inherited from cv::detail::FeaturesMatcher

Return

Name

Description

~FeaturesMatcher()

void

collectGarbage()

Frees unused memory allocated before if there is any.

bool

isThreadSafe()

void

operator()(
    const ImageFeatures & features1,
    const ImageFeatures & features2,
    MatchesInfo & matches_info )

void

operator()(
    const std::vector< ImageFeatures > & features,
    std::vector< MatchesInfo > & pairwise_matches,
    const cv::UMat & mask = cv::UMat() )

Performs images matching.

Additional Inherited Members#

Protected Member Functions inherited from cv::detail::FeaturesMatcher

Return

Name

Description

FeaturesMatcher(bool is_thread_safe = false)

void

match(
    const ImageFeatures & features1,
    const ImageFeatures & features2,
    MatchesInfo & matches_info )

This method must implement matching logic in order to make the wrappers detail::FeaturesMatcher::operator()_ work.

void

match(
    const std::vector< ImageFeatures > & features,
    std::vector< MatchesInfo > & pairwise_matches,
    const cv::UMat & mask = cv::UMat() )

This method implements logic to match features between arbitrary number of features. By default this checks every pair of inputs in the input, but the behaviour can be changed by subclasses.
Protected Attributes inherited from cv::detail::FeaturesMatcher

Return

Name

Description

bool

is_thread_safe_

Detailed Description#

Features matcher that adapts LightGlueMatcher (DescriptorMatcher) to the stitching pipeline’s FeaturesMatcher interface.

This matcher uses DNN-based LightGlue for feature matching, requiring ALIKED-style keypoints with spatial context for positional encoding.

See also

cv::detail::FeaturesMatcher cv::LightGlueMatcher

Examples
samples/cpp/stitching_detailed.cpp.

Constructor & Destructor Documentation#

LightGlueFeaturesMatcher()#

cv::detail::LightGlueFeaturesMatcher::LightGlueFeaturesMatcher(
Ptr< LightGlueMatcher > lgMatcher,
int num_matches_thresh1 = 6,
int num_matches_thresh2 = 6,
double matches_confidence_thresh = 3.0 )

Python:

cv.detail.LightGlueFeaturesMatcher(lgMatcher[, num_matches_thresh1[, num_matches_thresh2[, matches_confidence_thresh]]]) -> <detail_LightGlueFeaturesMatcher object>

Constructs a LightGlue features matcher.

Parameters

  • lgMatcher — LightGlueMatcher instance for DNN-based matching

  • num_matches_thresh1 — Minimum number of matches required for the 2D projective transform estimation used in the inliers classification step

  • num_matches_thresh2 — Minimum number of matches required for the 2D projective transform re-estimation on inliers

  • matches_confidence_thresh — Matching confidence threshold to take the match into account.

Member Function Documentation#

setScoreThreshold()#

void cv::detail::LightGlueFeaturesMatcher::setScoreThreshold(float thresh)

Python:

cv.detail.LightGlueFeaturesMatcher.setScoreThreshold(thresh)

Sets the LightGlue confidence threshold for filtering matches.

match()#

void cv::detail::LightGlueFeaturesMatcher::match(
const ImageFeatures & features1,
const ImageFeatures & features2,
MatchesInfo & matches_info )

This method must implement matching logic in order to make the wrappers detail::FeaturesMatcher::operator()_ work.

Parameters

  • features1 — first image features

  • features2 — second image features

  • matches_info — found matches

Member Data Documentation#

lg_score_thresh_#

float cv::detail::LightGlueFeaturesMatcher::lg_score_thresh_

lgMatcher_#

Ptr< LightGlueMatcher > cv::detail::LightGlueFeaturesMatcher::lgMatcher_

matches_confidence_thresh_#

double cv::detail::LightGlueFeaturesMatcher::matches_confidence_thresh_

num_matches_thresh1_#

int cv::detail::LightGlueFeaturesMatcher::num_matches_thresh1_

num_matches_thresh2_#

int cv::detail::LightGlueFeaturesMatcher::num_matches_thresh2_

Source file#

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