OpenCV  3.4.4
Open Source Computer Vision
Classes | Enumerations | Functions
Experimental 2D Features Algorithms

Classes

class  cv::xfeatures2d::AffineFeature2D
 Class implementing affine adaptation for key points. More...
 
class  cv::xfeatures2d::BoostDesc
 Class implementing BoostDesc (Learning Image Descriptors with Boosting), described in [180] and [181]. More...
 
class  cv::xfeatures2d::BriefDescriptorExtractor
 Class for computing BRIEF descriptors described in [30] . More...
 
class  cv::xfeatures2d::DAISY
 Class implementing DAISY descriptor, described in [187]. More...
 
class  cv::xfeatures2d::Elliptic_KeyPoint
 Elliptic region around an interest point. More...
 
class  cv::xfeatures2d::FREAK
 Class implementing the FREAK (Fast Retina Keypoint) keypoint descriptor, described in [4] . More...
 
class  cv::xfeatures2d::HarrisLaplaceFeatureDetector
 Class implementing the Harris-Laplace feature detector as described in [135]. More...
 
class  cv::xfeatures2d::LATCH
 
class  cv::xfeatures2d::LUCID
 Class implementing the locally uniform comparison image descriptor, described in [228]. More...
 
class  cv::xfeatures2d::MSDDetector
 Class implementing the MSD (Maximal Self-Dissimilarity) keypoint detector, described in [188]. More...
 
class  cv::xfeatures2d::PCTSignatures
 Class implementing PCT (position-color-texture) signature extraction as described in [104]. 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. More...
 
class  cv::xfeatures2d::PCTSignaturesSQFD
 Class implementing Signature Quadratic Form Distance (SQFD). More...
 
class  cv::xfeatures2d::StarDetector
 The class implements the keypoint detector introduced by [2], synonym of StarDetector. : More...
 
class  cv::xfeatures2d::VGG
 Class implementing VGG (Oxford Visual Geometry Group) descriptor trained end to end using "Descriptor Learning Using Convex Optimisation" (DLCO) aparatus described in [172]. More...
 

Enumerations

enum  {
  cv::xfeatures2d::FREAK::NB_SCALES = 64,
  cv::xfeatures2d::FREAK::NB_PAIRS = 512,
  cv::xfeatures2d::FREAK::NB_ORIENPAIRS = 45
}
 
enum  {
  cv::xfeatures2d::DAISY::NRM_NONE = 100,
  cv::xfeatures2d::DAISY::NRM_PARTIAL = 101,
  cv::xfeatures2d::DAISY::NRM_FULL = 102,
  cv::xfeatures2d::DAISY::NRM_SIFT = 103
}
 
enum  {
  cv::xfeatures2d::VGG::VGG_120 = 100,
  cv::xfeatures2d::VGG::VGG_80 = 101,
  cv::xfeatures2d::VGG::VGG_64 = 102,
  cv::xfeatures2d::VGG::VGG_48 = 103
}
 
enum  {
  cv::xfeatures2d::BoostDesc::BGM = 100,
  cv::xfeatures2d::BoostDesc::BGM_HARD = 101,
  cv::xfeatures2d::BoostDesc::BGM_BILINEAR = 102,
  cv::xfeatures2d::BoostDesc::LBGM = 200,
  cv::xfeatures2d::BoostDesc::BINBOOST_64 = 300,
  cv::xfeatures2d::BoostDesc::BINBOOST_128 = 301,
  cv::xfeatures2d::BoostDesc::BINBOOST_256 = 302
}
 
enum  cv::xfeatures2d::PCTSignatures::DistanceFunction {
  cv::xfeatures2d::PCTSignatures::L0_25,
  cv::xfeatures2d::PCTSignatures::L0_5,
  cv::xfeatures2d::PCTSignatures::L1,
  cv::xfeatures2d::PCTSignatures::L2,
  cv::xfeatures2d::PCTSignatures::L2SQUARED,
  cv::xfeatures2d::PCTSignatures::L5,
  cv::xfeatures2d::PCTSignatures::L_INFINITY
}
 Lp distance function selector. More...
 
enum  cv::xfeatures2d::PCTSignatures::PointDistribution {
  cv::xfeatures2d::PCTSignatures::UNIFORM,
  cv::xfeatures2d::PCTSignatures::REGULAR,
  cv::xfeatures2d::PCTSignatures::NORMAL
}
 Point distributions supported by random point generator. More...
 
enum  cv::xfeatures2d::PCTSignatures::SimilarityFunction {
  cv::xfeatures2d::PCTSignatures::MINUS,
  cv::xfeatures2d::PCTSignatures::GAUSSIAN,
  cv::xfeatures2d::PCTSignatures::HEURISTIC
}
 Similarity function selector. More...
 

Functions

void cv::xfeatures2d::FASTForPointSet (InputArray image, std::vector< KeyPoint > &keypoints, int threshold, bool nonmaxSuppression=true, int type=FastFeatureDetector::TYPE_9_16)
 Estimates cornerness for prespecified KeyPoints using the FAST algorithm. More...
 

Detailed Description

This section describes experimental algorithms for 2d feature detection.

Enumeration Type Documentation

anonymous enum
Enumerator
NB_SCALES 
NB_PAIRS 
NB_ORIENPAIRS 
anonymous enum
Enumerator
NRM_NONE 
NRM_PARTIAL 
NRM_FULL 
NRM_SIFT 
anonymous enum
Enumerator
VGG_120 
VGG_80 
VGG_64 
VGG_48 
anonymous enum
Enumerator
BGM 
BGM_HARD 
BGM_BILINEAR 
LBGM 
BINBOOST_64 
BINBOOST_128 
BINBOOST_256 

Lp distance function selector.

Enumerator
L0_25 
L0_5 
L1 
L2 
L2SQUARED 
L5 
L_INFINITY 

Point distributions supported by random point generator.

Enumerator
UNIFORM 

Generate numbers uniformly.

REGULAR 

Generate points in a regular grid.

NORMAL 

Generate points with normal (gaussian) distribution.

Similarity function selector.

See also
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. [11]
Note
For selected distance function:

\[ d(c_i, c_j) \]

and parameter:

\[ \alpha \]

Enumerator
MINUS 

\[ -d(c_i, c_j) \]

GAUSSIAN 

\[ e^{ -\alpha * d^2(c_i, c_j)} \]

HEURISTIC 

\[ \frac{1}{\alpha + d(c_i, c_j)} \]

Function Documentation

void cv::xfeatures2d::FASTForPointSet ( InputArray  image,
std::vector< KeyPoint > &  keypoints,
int  threshold,
bool  nonmaxSuppression = true,
int  type = FastFeatureDetector::TYPE_9_16 
)

Estimates cornerness for prespecified KeyPoints using the FAST algorithm.

Parameters
imagegrayscale image where keypoints (corners) are detected.
keypointskeypoints which should be tested to fit the FAST criteria. Keypoints not beeing detected as corners are removed.
thresholdthreshold on difference between intensity of the central pixel and pixels of a circle around this pixel.
nonmaxSuppressionif true, non-maximum suppression is applied to detected corners (keypoints).
typeone of the three neighborhoods as defined in the paper: FastFeatureDetector::TYPE_9_16, FastFeatureDetector::TYPE_7_12, FastFeatureDetector::TYPE_5_8

Detects corners using the FAST algorithm by [162] .