1 /*M///////////////////////////////////////////////////////////////////////////////////////
3 // IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
5 // By downloading, copying, installing or using the software you agree to this license.
6 // If you do not agree to this license, do not download, install,
7 // copy or use the software.
11 // For Open Source Computer Vision Library
13 // Copyright (C) 2000-2008, Intel Corporation, all rights reserved.
14 // Copyright (C) 2009, Willow Garage Inc., all rights reserved.
15 // Third party copyrights are property of their respective owners.
17 // Redistribution and use in source and binary forms, with or without modification,
18 // are permitted provided that the following conditions are met:
20 // * Redistribution's of source code must retain the above copyright notice,
21 // this list of conditions and the following disclaimer.
23 // * Redistribution's in binary form must reproduce the above copyright notice,
24 // this list of conditions and the following disclaimer in the documentation
25 // and/or other materials provided with the distribution.
27 // * The name of the copyright holders may not be used to endorse or promote products
28 // derived from this software without specific prior written permission.
30 // This software is provided by the copyright holders and contributors "as is" and
31 // any express or implied warranties, including, but not limited to, the implied
32 // warranties of merchantability and fitness for a particular purpose are disclaimed.
33 // In no event shall the Intel Corporation or contributors be liable for any direct,
34 // indirect, incidental, special, exemplary, or consequential damages
35 // (including, but not limited to, procurement of substitute goods or services;
36 // loss of use, data, or profits; or business interruption) however caused
37 // and on any theory of liability, whether in contract, strict liability,
38 // or tort (including negligence or otherwise) arising in any way out of
39 // the use of this software, even if advised of the possibility of such damage.
43 #ifndef __OPENCV_FEATURES_2D_HPP__
44 #define __OPENCV_FEATURES_2D_HPP__
46 #include "opencv2/core.hpp"
47 #include "opencv2/flann/miniflann.hpp"
52 CV_EXPORTS bool initModule_features2d();
54 // //! writes vector of keypoints to the file storage
55 // CV_EXPORTS void write(FileStorage& fs, const String& name, const std::vector<KeyPoint>& keypoints);
56 // //! reads vector of keypoints from the specified file storage node
57 // CV_EXPORTS void read(const FileNode& node, CV_OUT std::vector<KeyPoint>& keypoints);
60 * A class filters a vector of keypoints.
61 * Because now it is difficult to provide a convenient interface for all usage scenarios of the keypoints filter class,
62 * it has only several needed by now static methods.
64 class CV_EXPORTS KeyPointsFilter
70 * Remove keypoints within borderPixels of an image edge.
72 static void runByImageBorder( std::vector<KeyPoint>& keypoints, Size imageSize, int borderSize );
74 * Remove keypoints of sizes out of range.
76 static void runByKeypointSize( std::vector<KeyPoint>& keypoints, float minSize,
77 float maxSize=FLT_MAX );
79 * Remove keypoints from some image by mask for pixels of this image.
81 static void runByPixelsMask( std::vector<KeyPoint>& keypoints, const Mat& mask );
83 * Remove duplicated keypoints.
85 static void removeDuplicated( std::vector<KeyPoint>& keypoints );
88 * Retain the specified number of the best keypoints (according to the response)
90 static void retainBest( std::vector<KeyPoint>& keypoints, int npoints );
94 /************************************ Base Classes ************************************/
97 * Abstract base class for 2D image feature detectors.
99 class CV_EXPORTS_W FeatureDetector : public virtual Algorithm
102 virtual ~FeatureDetector();
105 * Detect keypoints in an image.
107 * keypoints The detected keypoints.
108 * mask Mask specifying where to look for keypoints (optional). Must be a char
109 * matrix with non-zero values in the region of interest.
111 CV_WRAP void detect( InputArray image, CV_OUT std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
114 * Detect keypoints in an image set.
115 * images Image collection.
116 * keypoints Collection of keypoints detected in an input images. keypoints[i] is a set of keypoints detected in an images[i].
117 * masks Masks for image set. masks[i] is a mask for images[i].
119 void detect( InputArrayOfArrays images, std::vector<std::vector<KeyPoint> >& keypoints, InputArrayOfArrays masks=noArray() ) const;
121 // Return true if detector object is empty
122 CV_WRAP virtual bool empty() const;
124 // Create feature detector by detector name.
125 CV_WRAP static Ptr<FeatureDetector> create( const String& detectorType );
128 virtual void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const = 0;
131 * Remove keypoints that are not in the mask.
132 * Helper function, useful when wrapping a library call for keypoint detection that
133 * does not support a mask argument.
135 static void removeInvalidPoints( const Mat & mask, std::vector<KeyPoint>& keypoints );
140 * Abstract base class for computing descriptors for image keypoints.
142 * In this interface we assume a keypoint descriptor can be represented as a
143 * dense, fixed-dimensional vector of some basic type. Most descriptors used
144 * in practice follow this pattern, as it makes it very easy to compute
145 * distances between descriptors. Therefore we represent a collection of
146 * descriptors as a Mat, where each row is one keypoint descriptor.
148 class CV_EXPORTS_W DescriptorExtractor : public virtual Algorithm
151 virtual ~DescriptorExtractor();
154 * Compute the descriptors for a set of keypoints in an image.
156 * keypoints The input keypoints. Keypoints for which a descriptor cannot be computed are removed.
157 * descriptors Copmputed descriptors. Row i is the descriptor for keypoint i.
159 CV_WRAP void compute( InputArray image, CV_OUT CV_IN_OUT std::vector<KeyPoint>& keypoints, OutputArray descriptors ) const;
162 * Compute the descriptors for a keypoints collection detected in image collection.
163 * images Image collection.
164 * keypoints Input keypoints collection. keypoints[i] is keypoints detected in images[i].
165 * Keypoints for which a descriptor cannot be computed are removed.
166 * descriptors Descriptor collection. descriptors[i] are descriptors computed for set keypoints[i].
168 void compute( InputArrayOfArrays images, std::vector<std::vector<KeyPoint> >& keypoints, OutputArrayOfArrays descriptors ) const;
170 CV_WRAP virtual int descriptorSize() const = 0;
171 CV_WRAP virtual int descriptorType() const = 0;
172 CV_WRAP virtual int defaultNorm() const = 0;
174 CV_WRAP virtual bool empty() const;
176 CV_WRAP static Ptr<DescriptorExtractor> create( const String& descriptorExtractorType );
179 virtual void computeImpl( InputArray image, std::vector<KeyPoint>& keypoints, OutputArray descriptors ) const = 0;
182 * Remove keypoints within borderPixels of an image edge.
184 static void removeBorderKeypoints( std::vector<KeyPoint>& keypoints,
185 Size imageSize, int borderSize );
191 * Abstract base class for simultaneous 2D feature detection descriptor extraction.
193 class CV_EXPORTS_W Feature2D : public FeatureDetector, public DescriptorExtractor
197 * Detect keypoints in an image.
199 * keypoints The detected keypoints.
200 * mask Mask specifying where to look for keypoints (optional). Must be a char
201 * matrix with non-zero values in the region of interest.
202 * useProvidedKeypoints If true, the method will skip the detection phase and will compute
203 * descriptors for the provided keypoints
205 CV_WRAP_AS(detectAndCompute) virtual void operator()( InputArray image, InputArray mask,
206 CV_OUT std::vector<KeyPoint>& keypoints,
207 OutputArray descriptors,
208 bool useProvidedKeypoints=false ) const = 0;
210 CV_WRAP void compute( InputArray image, CV_OUT CV_IN_OUT std::vector<KeyPoint>& keypoints, OutputArray descriptors ) const;
212 // Create feature detector and descriptor extractor by name.
213 CV_WRAP static Ptr<Feature2D> create( const String& name );
219 class CV_EXPORTS_W BRISK : public Feature2D
222 CV_WRAP explicit BRISK(int thresh=30, int octaves=3, float patternScale=1.0f);
226 // returns the descriptor size in bytes
227 int descriptorSize() const;
228 // returns the descriptor type
229 int descriptorType() const;
230 // returns the default norm type
231 int defaultNorm() const;
233 // Compute the BRISK features on an image
234 void operator()(InputArray image, InputArray mask, std::vector<KeyPoint>& keypoints) const;
236 // Compute the BRISK features and descriptors on an image
237 void operator()( InputArray image, InputArray mask, std::vector<KeyPoint>& keypoints,
238 OutputArray descriptors, bool useProvidedKeypoints=false ) const;
240 AlgorithmInfo* info() const;
243 CV_WRAP explicit BRISK(std::vector<float> &radiusList, std::vector<int> &numberList,
244 float dMax=5.85f, float dMin=8.2f, std::vector<int> indexChange=std::vector<int>());
246 // call this to generate the kernel:
247 // circle of radius r (pixels), with n points;
248 // short pairings with dMax, long pairings with dMin
249 CV_WRAP void generateKernel(std::vector<float> &radiusList,
250 std::vector<int> &numberList, float dMax=5.85f, float dMin=8.2f,
251 std::vector<int> indexChange=std::vector<int>());
255 void computeImpl( InputArray image, std::vector<KeyPoint>& keypoints, OutputArray descriptors ) const;
256 void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
258 void computeKeypointsNoOrientation(InputArray image, InputArray mask, std::vector<KeyPoint>& keypoints) const;
259 void computeDescriptorsAndOrOrientation(InputArray image, InputArray mask, std::vector<KeyPoint>& keypoints,
260 OutputArray descriptors, bool doDescriptors, bool doOrientation,
261 bool useProvidedKeypoints) const;
263 // Feature parameters
264 CV_PROP_RW int threshold;
265 CV_PROP_RW int octaves;
267 // some helper structures for the Brisk pattern representation
268 struct BriskPatternPoint{
269 float x; // x coordinate relative to center
270 float y; // x coordinate relative to center
271 float sigma; // Gaussian smoothing sigma
273 struct BriskShortPair{
274 unsigned int i; // index of the first pattern point
275 unsigned int j; // index of other pattern point
277 struct BriskLongPair{
278 unsigned int i; // index of the first pattern point
279 unsigned int j; // index of other pattern point
280 int weighted_dx; // 1024.0/dx
281 int weighted_dy; // 1024.0/dy
283 inline int smoothedIntensity(const cv::Mat& image,
284 const cv::Mat& integral,const float key_x,
285 const float key_y, const unsigned int scale,
286 const unsigned int rot, const unsigned int point) const;
287 // pattern properties
288 BriskPatternPoint* patternPoints_; //[i][rotation][scale]
289 unsigned int points_; // total number of collocation points
290 float* scaleList_; // lists the scaling per scale index [scale]
291 unsigned int* sizeList_; // lists the total pattern size per scale index [scale]
292 static const unsigned int scales_; // scales discretization
293 static const float scalerange_; // span of sizes 40->4 Octaves - else, this needs to be adjusted...
294 static const unsigned int n_rot_; // discretization of the rotation look-up
297 int strings_; // number of uchars the descriptor consists of
298 float dMax_; // short pair maximum distance
299 float dMin_; // long pair maximum distance
300 BriskShortPair* shortPairs_; // d<_dMax
301 BriskLongPair* longPairs_; // d>_dMin
302 unsigned int noShortPairs_; // number of shortParis
303 unsigned int noLongPairs_; // number of longParis
306 static const float basicSize_;
313 class CV_EXPORTS_W ORB : public Feature2D
316 // the size of the signature in bytes
317 enum { kBytes = 32, HARRIS_SCORE=0, FAST_SCORE=1 };
319 CV_WRAP explicit ORB(int nfeatures = 500, float scaleFactor = 1.2f, int nlevels = 8, int edgeThreshold = 31,
320 int firstLevel = 0, int WTA_K=2, int scoreType=ORB::HARRIS_SCORE, int patchSize=31 );
322 // returns the descriptor size in bytes
323 int descriptorSize() const;
324 // returns the descriptor type
325 int descriptorType() const;
326 // returns the default norm type
327 int defaultNorm() const;
329 // Compute the ORB features and descriptors on an image
330 void operator()(InputArray image, InputArray mask, std::vector<KeyPoint>& keypoints) const;
332 // Compute the ORB features and descriptors on an image
333 void operator()( InputArray image, InputArray mask, std::vector<KeyPoint>& keypoints,
334 OutputArray descriptors, bool useProvidedKeypoints=false ) const;
336 AlgorithmInfo* info() const;
340 void computeImpl( InputArray image, std::vector<KeyPoint>& keypoints, OutputArray descriptors ) const;
341 void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
343 CV_PROP_RW int nfeatures;
344 CV_PROP_RW double scaleFactor;
345 CV_PROP_RW int nlevels;
346 CV_PROP_RW int edgeThreshold;
347 CV_PROP_RW int firstLevel;
348 CV_PROP_RW int WTA_K;
349 CV_PROP_RW int scoreType;
350 CV_PROP_RW int patchSize;
353 typedef ORB OrbFeatureDetector;
354 typedef ORB OrbDescriptorExtractor;
359 class CV_EXPORTS FREAK : public DescriptorExtractor
363 * @param orientationNormalized enable orientation normalization
364 * @param scaleNormalized enable scale normalization
365 * @param patternScale scaling of the description pattern
366 * @param nbOctave number of octaves covered by the detected keypoints
367 * @param selectedPairs (optional) user defined selected pairs
369 explicit FREAK( bool orientationNormalized = true,
370 bool scaleNormalized = true,
371 float patternScale = 22.0f,
373 const std::vector<int>& selectedPairs = std::vector<int>());
374 FREAK( const FREAK& rhs );
375 FREAK& operator=( const FREAK& );
379 /** returns the descriptor length in bytes */
380 virtual int descriptorSize() const;
382 /** returns the descriptor type */
383 virtual int descriptorType() const;
385 /** returns the default norm type */
386 virtual int defaultNorm() const;
388 /** select the 512 "best description pairs"
389 * @param images grayscale images set
390 * @param keypoints set of detected keypoints
391 * @param corrThresh correlation threshold
392 * @param verbose print construction information
393 * @return list of best pair indexes
395 std::vector<int> selectPairs( const std::vector<Mat>& images, std::vector<std::vector<KeyPoint> >& keypoints,
396 const double corrThresh = 0.7, bool verbose = true );
398 AlgorithmInfo* info() const;
402 NB_SCALES = 64, NB_PAIRS = 512, NB_ORIENPAIRS = 45
406 virtual void computeImpl( InputArray image, std::vector<KeyPoint>& keypoints, OutputArray descriptors ) const;
409 template <typename imgType, typename iiType>
410 imgType meanIntensity( InputArray image, InputArray integral, const float kp_x, const float kp_y,
411 const unsigned int scale, const unsigned int rot, const unsigned int point ) const;
413 template <typename srcMatType, typename iiMatType>
414 void computeDescriptors( InputArray image, std::vector<KeyPoint>& keypoints, OutputArray descriptors ) const;
416 template <typename srcMatType>
417 void extractDescriptor(srcMatType *pointsValue, void ** ptr) const;
419 bool orientationNormalized; //true if the orientation is normalized, false otherwise
420 bool scaleNormalized; //true if the scale is normalized, false otherwise
421 double patternScale; //scaling of the pattern
422 int nOctaves; //number of octaves
423 bool extAll; // true if all pairs need to be extracted for pairs selection
425 double patternScale0;
427 std::vector<int> selectedPairs0;
431 float x; // x coordinate relative to center
432 float y; // x coordinate relative to center
433 float sigma; // Gaussian smoothing sigma
436 struct DescriptionPair
438 uchar i; // index of the first point
439 uchar j; // index of the second point
442 struct OrientationPair
444 uchar i; // index of the first point
445 uchar j; // index of the second point
446 int weight_dx; // dx/(norm_sq))*4096
447 int weight_dy; // dy/(norm_sq))*4096
450 std::vector<PatternPoint> patternLookup; // look-up table for the pattern points (position+sigma of all points at all scales and orientation)
451 int patternSizes[NB_SCALES]; // size of the pattern at a specific scale (used to check if a point is within image boundaries)
452 DescriptionPair descriptionPairs[NB_PAIRS];
453 OrientationPair orientationPairs[NB_ORIENPAIRS];
458 Maximal Stable Extremal Regions class.
460 The class implements MSER algorithm introduced by J. Matas.
461 Unlike SIFT, SURF and many other detectors in OpenCV, this is salient region detector,
462 not the salient point detector.
464 It returns the regions, each of those is encoded as a contour.
466 class CV_EXPORTS_W MSER : public FeatureDetector
469 //! the full constructor
470 CV_WRAP explicit MSER( int _delta=5, int _min_area=60, int _max_area=14400,
471 double _max_variation=0.25, double _min_diversity=.2,
472 int _max_evolution=200, double _area_threshold=1.01,
473 double _min_margin=0.003, int _edge_blur_size=5 );
475 //! the operator that extracts the MSERs from the image or the specific part of it
476 CV_WRAP_AS(detect) void operator()( InputArray image, CV_OUT std::vector<std::vector<Point> >& msers,
477 InputArray mask=noArray() ) const;
478 AlgorithmInfo* info() const;
481 void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
489 double areaThreshold;
494 typedef MSER MserFeatureDetector;
499 The class implements the keypoint detector introduced by K. Konolige.
501 class CV_EXPORTS_W StarDetector : public FeatureDetector
504 //! the full constructor
505 CV_WRAP StarDetector(int _maxSize=45, int _responseThreshold=30,
506 int _lineThresholdProjected=10,
507 int _lineThresholdBinarized=8,
508 int _suppressNonmaxSize=5);
510 //! finds the keypoints in the image
511 CV_WRAP_AS(detect) void operator()(const Mat& image,
512 CV_OUT std::vector<KeyPoint>& keypoints) const;
514 AlgorithmInfo* info() const;
517 void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
520 int responseThreshold;
521 int lineThresholdProjected;
522 int lineThresholdBinarized;
523 int suppressNonmaxSize;
526 //! detects corners using FAST algorithm by E. Rosten
527 CV_EXPORTS void FAST( InputArray image, CV_OUT std::vector<KeyPoint>& keypoints,
528 int threshold, bool nonmaxSuppression=true );
530 CV_EXPORTS void FAST( InputArray image, CV_OUT std::vector<KeyPoint>& keypoints,
531 int threshold, bool nonmaxSuppression, int type );
533 class CV_EXPORTS_W FastFeatureDetector : public FeatureDetector
538 TYPE_5_8 = 0, TYPE_7_12 = 1, TYPE_9_16 = 2
541 CV_WRAP FastFeatureDetector( int threshold=10, bool nonmaxSuppression=true);
542 CV_WRAP FastFeatureDetector( int threshold, bool nonmaxSuppression, int type);
543 AlgorithmInfo* info() const;
546 virtual void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
549 bool nonmaxSuppression;
554 class CV_EXPORTS_W GFTTDetector : public FeatureDetector
557 CV_WRAP GFTTDetector( int maxCorners=1000, double qualityLevel=0.01, double minDistance=1,
558 int blockSize=3, bool useHarrisDetector=false, double k=0.04 );
559 AlgorithmInfo* info() const;
562 virtual void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
568 bool useHarrisDetector;
572 typedef GFTTDetector GoodFeaturesToTrackDetector;
573 typedef StarDetector StarFeatureDetector;
575 class CV_EXPORTS_W SimpleBlobDetector : public FeatureDetector
578 struct CV_EXPORTS_W_SIMPLE Params
581 CV_PROP_RW float thresholdStep;
582 CV_PROP_RW float minThreshold;
583 CV_PROP_RW float maxThreshold;
584 CV_PROP_RW size_t minRepeatability;
585 CV_PROP_RW float minDistBetweenBlobs;
587 CV_PROP_RW bool filterByColor;
588 CV_PROP_RW uchar blobColor;
590 CV_PROP_RW bool filterByArea;
591 CV_PROP_RW float minArea, maxArea;
593 CV_PROP_RW bool filterByCircularity;
594 CV_PROP_RW float minCircularity, maxCircularity;
596 CV_PROP_RW bool filterByInertia;
597 CV_PROP_RW float minInertiaRatio, maxInertiaRatio;
599 CV_PROP_RW bool filterByConvexity;
600 CV_PROP_RW float minConvexity, maxConvexity;
602 void read( const FileNode& fn );
603 void write( FileStorage& fs ) const;
606 CV_WRAP SimpleBlobDetector(const SimpleBlobDetector::Params ¶meters = SimpleBlobDetector::Params());
608 virtual void read( const FileNode& fn );
609 virtual void write( FileStorage& fs ) const;
612 struct CV_EXPORTS Center
619 virtual void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
620 virtual void findBlobs(InputArray image, InputArray binaryImage, std::vector<Center> ¢ers) const;
623 AlgorithmInfo* info() const;
627 class CV_EXPORTS_W DenseFeatureDetector : public FeatureDetector
630 CV_WRAP explicit DenseFeatureDetector( float initFeatureScale=1.f, int featureScaleLevels=1,
631 float featureScaleMul=0.1f,
632 int initXyStep=6, int initImgBound=0,
633 bool varyXyStepWithScale=true,
634 bool varyImgBoundWithScale=false );
635 AlgorithmInfo* info() const;
638 virtual void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
640 double initFeatureScale;
641 int featureScaleLevels;
642 double featureScaleMul;
647 bool varyXyStepWithScale;
648 bool varyImgBoundWithScale;
652 * Adapts a detector to partition the source image into a grid and detect
653 * points in each cell.
655 class CV_EXPORTS_W GridAdaptedFeatureDetector : public FeatureDetector
659 * detector Detector that will be adapted.
660 * maxTotalKeypoints Maximum count of keypoints detected on the image. Only the strongest keypoints
662 * gridRows Grid rows count.
663 * gridCols Grid column count.
665 CV_WRAP GridAdaptedFeatureDetector( const Ptr<FeatureDetector>& detector=Ptr<FeatureDetector>(),
666 int maxTotalKeypoints=1000,
667 int gridRows=4, int gridCols=4 );
669 // TODO implement read/write
670 virtual bool empty() const;
672 AlgorithmInfo* info() const;
675 virtual void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
677 Ptr<FeatureDetector> detector;
678 int maxTotalKeypoints;
684 * Adapts a detector to detect points over multiple levels of a Gaussian
685 * pyramid. Useful for detectors that are not inherently scaled.
687 class CV_EXPORTS_W PyramidAdaptedFeatureDetector : public FeatureDetector
690 // maxLevel - The 0-based index of the last pyramid layer
691 CV_WRAP PyramidAdaptedFeatureDetector( const Ptr<FeatureDetector>& detector, int maxLevel=2 );
693 // TODO implement read/write
694 virtual bool empty() const;
697 virtual void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
699 Ptr<FeatureDetector> detector;
703 /** \brief A feature detector parameter adjuster, this is used by the DynamicAdaptedFeatureDetector
704 * and is a wrapper for FeatureDetector that allow them to be adjusted after a detection
706 class CV_EXPORTS AdjusterAdapter: public FeatureDetector
709 /** pure virtual interface
711 virtual ~AdjusterAdapter() {}
712 /** too few features were detected so, adjust the detector params accordingly
713 * \param min the minimum number of desired features
714 * \param n_detected the number previously detected
716 virtual void tooFew(int min, int n_detected) = 0;
717 /** too many features were detected so, adjust the detector params accordingly
718 * \param max the maximum number of desired features
719 * \param n_detected the number previously detected
721 virtual void tooMany(int max, int n_detected) = 0;
722 /** are params maxed out or still valid?
723 * \return false if the parameters can't be adjusted any more
725 virtual bool good() const = 0;
727 virtual Ptr<AdjusterAdapter> clone() const = 0;
729 static Ptr<AdjusterAdapter> create( const String& detectorType );
731 /** \brief an adaptively adjusting detector that iteratively detects until the desired number
732 * of features are detected.
733 * Beware that this is not thread safe - as the adjustment of parameters breaks the const
734 * of the detection routine...
735 * /TODO Make this const correct and thread safe
738 //will create a detector that attempts to find 100 - 110 FAST Keypoints, and will at most run
739 //FAST feature detection 10 times until that number of keypoints are found
740 Ptr<FeatureDetector> detector(new DynamicAdaptedFeatureDetector(new FastAdjuster(20,true),100, 110, 10));
743 class CV_EXPORTS DynamicAdaptedFeatureDetector: public FeatureDetector
747 /** \param adjuster an AdjusterAdapter that will do the detection and parameter adjustment
748 * \param max_features the maximum desired number of features
749 * \param max_iters the maximum number of times to try to adjust the feature detector params
750 * for the FastAdjuster this can be high, but with Star or Surf this can get time consuming
751 * \param min_features the minimum desired features
753 DynamicAdaptedFeatureDetector( const Ptr<AdjusterAdapter>& adjuster, int min_features=400, int max_features=500, int max_iters=5 );
755 virtual bool empty() const;
758 virtual void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
761 DynamicAdaptedFeatureDetector& operator=(const DynamicAdaptedFeatureDetector&);
762 DynamicAdaptedFeatureDetector(const DynamicAdaptedFeatureDetector&);
765 int min_features_, max_features_;
766 const Ptr<AdjusterAdapter> adjuster_;
769 /**\brief an adjust for the FAST detector. This will basically decrement or increment the
772 class CV_EXPORTS FastAdjuster: public AdjusterAdapter
775 /**\param init_thresh the initial threshold to start with, default = 20
776 * \param nonmax whether to use non max or not for fast feature detection
778 FastAdjuster(int init_thresh=20, bool nonmax=true, int min_thresh=1, int max_thresh=200);
780 virtual void tooFew(int minv, int n_detected);
781 virtual void tooMany(int maxv, int n_detected);
782 virtual bool good() const;
784 virtual Ptr<AdjusterAdapter> clone() const;
787 virtual void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
791 int init_thresh_, min_thresh_, max_thresh_;
795 /** An adjuster for StarFeatureDetector, this one adjusts the responseThreshold for now
796 * TODO find a faster way to converge the parameters for Star - use CvStarDetectorParams
798 class CV_EXPORTS StarAdjuster: public AdjusterAdapter
801 StarAdjuster(double initial_thresh=30.0, double min_thresh=2., double max_thresh=200.);
803 virtual void tooFew(int minv, int n_detected);
804 virtual void tooMany(int maxv, int n_detected);
805 virtual bool good() const;
807 virtual Ptr<AdjusterAdapter> clone() const;
810 virtual void detectImpl(InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
812 double thresh_, init_thresh_, min_thresh_, max_thresh_;
815 class CV_EXPORTS SurfAdjuster: public AdjusterAdapter
818 SurfAdjuster( double initial_thresh=400.f, double min_thresh=2, double max_thresh=1000 );
820 virtual void tooFew(int minv, int n_detected);
821 virtual void tooMany(int maxv, int n_detected);
822 virtual bool good() const;
824 virtual Ptr<AdjusterAdapter> clone() const;
827 virtual void detectImpl( InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask=noArray() ) const;
829 double thresh_, init_thresh_, min_thresh_, max_thresh_;
832 CV_EXPORTS Mat windowedMatchingMask( const std::vector<KeyPoint>& keypoints1, const std::vector<KeyPoint>& keypoints2,
833 float maxDeltaX, float maxDeltaY );
838 * OpponentColorDescriptorExtractor
840 * Adapts a descriptor extractor to compute descriptors in Opponent Color Space
841 * (refer to van de Sande et al., CGIV 2008 "Color Descriptors for Object Category Recognition").
842 * Input RGB image is transformed in Opponent Color Space. Then unadapted descriptor extractor
843 * (set in constructor) computes descriptors on each of the three channel and concatenate
844 * them into a single color descriptor.
846 class CV_EXPORTS OpponentColorDescriptorExtractor : public DescriptorExtractor
849 OpponentColorDescriptorExtractor( const Ptr<DescriptorExtractor>& descriptorExtractor );
851 virtual void read( const FileNode& );
852 virtual void write( FileStorage& ) const;
854 virtual int descriptorSize() const;
855 virtual int descriptorType() const;
856 virtual int defaultNorm() const;
858 virtual bool empty() const;
861 virtual void computeImpl( InputArray image, std::vector<KeyPoint>& keypoints, OutputArray descriptors ) const;
863 Ptr<DescriptorExtractor> descriptorExtractor;
869 class CV_EXPORTS BriefDescriptorExtractor : public DescriptorExtractor
872 static const int PATCH_SIZE = 48;
873 static const int KERNEL_SIZE = 9;
875 // bytes is a length of descriptor in bytes. It can be equal 16, 32 or 64 bytes.
876 BriefDescriptorExtractor( int bytes = 32 );
878 virtual void read( const FileNode& );
879 virtual void write( FileStorage& ) const;
881 virtual int descriptorSize() const;
882 virtual int descriptorType() const;
883 virtual int defaultNorm() const;
885 /// @todo read and write for brief
887 AlgorithmInfo* info() const;
890 virtual void computeImpl(InputArray image, std::vector<KeyPoint>& keypoints, OutputArray descriptors) const;
892 typedef void(*PixelTestFn)(InputArray, const std::vector<KeyPoint>&, OutputArray);
895 PixelTestFn test_fn_;
901 class CV_EXPORTS_W KAZE : public Feature2D
905 CV_WRAP explicit KAZE(bool extended, bool upright);
909 // returns the descriptor size in bytes
910 int descriptorSize() const;
911 // returns the descriptor type
912 int descriptorType() const;
913 // returns the default norm type
914 int defaultNorm() const;
916 AlgorithmInfo* info() const;
918 // Compute the KAZE features on an image
919 void operator()(InputArray image, InputArray mask, std::vector<KeyPoint>& keypoints) const;
921 // Compute the KAZE features and descriptors on an image
922 void operator()(InputArray image, InputArray mask, std::vector<KeyPoint>& keypoints,
923 OutputArray descriptors, bool useProvidedKeypoints = false) const;
926 void detectImpl(InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask) const;
927 void computeImpl(InputArray image, std::vector<KeyPoint>& keypoints, OutputArray descriptors) const;
929 CV_PROP bool extended;
930 CV_PROP bool upright;
936 class CV_EXPORTS_W AKAZE : public Feature2D
939 /// AKAZE Descriptor Type
940 enum DESCRIPTOR_TYPE {
941 DESCRIPTOR_KAZE_UPRIGHT = 2, ///< Upright descriptors, not invariant to rotation
943 DESCRIPTOR_MLDB_UPRIGHT = 4, ///< Upright descriptors, not invariant to rotation
948 explicit AKAZE(DESCRIPTOR_TYPE descriptor_type, int descriptor_size = 0, int descriptor_channels = 3);
952 // returns the descriptor size in bytes
953 int descriptorSize() const;
954 // returns the descriptor type
955 int descriptorType() const;
956 // returns the default norm type
957 int defaultNorm() const;
959 // Compute the AKAZE features on an image
960 void operator()(InputArray image, InputArray mask, std::vector<KeyPoint>& keypoints) const;
962 // Compute the AKAZE features and descriptors on an image
963 void operator()(InputArray image, InputArray mask, std::vector<KeyPoint>& keypoints,
964 OutputArray descriptors, bool useProvidedKeypoints = false) const;
966 AlgorithmInfo* info() const;
970 void computeImpl(InputArray image, std::vector<KeyPoint>& keypoints, OutputArray descriptors) const;
971 void detectImpl(InputArray image, std::vector<KeyPoint>& keypoints, InputArray mask = noArray()) const;
973 CV_PROP int descriptor;
974 CV_PROP int descriptor_channels;
975 CV_PROP int descriptor_size;
978 /****************************************************************************************\
980 \****************************************************************************************/
983 struct CV_EXPORTS Accumulator
988 template<> struct Accumulator<unsigned char> { typedef float Type; };
989 template<> struct Accumulator<unsigned short> { typedef float Type; };
990 template<> struct Accumulator<char> { typedef float Type; };
991 template<> struct Accumulator<short> { typedef float Type; };
994 * Squared Euclidean distance functor
997 struct CV_EXPORTS SL2
999 enum { normType = NORM_L2SQR };
1000 typedef T ValueType;
1001 typedef typename Accumulator<T>::Type ResultType;
1003 ResultType operator()( const T* a, const T* b, int size ) const
1005 return normL2Sqr<ValueType, ResultType>(a, b, size);
1010 * Euclidean distance functor
1013 struct CV_EXPORTS L2
1015 enum { normType = NORM_L2 };
1016 typedef T ValueType;
1017 typedef typename Accumulator<T>::Type ResultType;
1019 ResultType operator()( const T* a, const T* b, int size ) const
1021 return (ResultType)std::sqrt((double)normL2Sqr<ValueType, ResultType>(a, b, size));
1026 * Manhattan distance (city block distance) functor
1029 struct CV_EXPORTS L1
1031 enum { normType = NORM_L1 };
1032 typedef T ValueType;
1033 typedef typename Accumulator<T>::Type ResultType;
1035 ResultType operator()( const T* a, const T* b, int size ) const
1037 return normL1<ValueType, ResultType>(a, b, size);
1042 * Hamming distance functor - counts the bit differences between two strings - useful for the Brief descriptor
1043 * bit count of A exclusive XOR'ed with B
1045 struct CV_EXPORTS Hamming
1047 enum { normType = NORM_HAMMING };
1048 typedef unsigned char ValueType;
1049 typedef int ResultType;
1051 /** this will count the bits in a ^ b
1053 ResultType operator()( const unsigned char* a, const unsigned char* b, int size ) const
1055 return normHamming(a, b, size);
1059 typedef Hamming HammingLUT;
1061 template<int cellsize> struct HammingMultilevel
1063 enum { normType = NORM_HAMMING + (cellsize>1) };
1064 typedef unsigned char ValueType;
1065 typedef int ResultType;
1067 ResultType operator()( const unsigned char* a, const unsigned char* b, int size ) const
1069 return normHamming(a, b, size, cellsize);
1073 /****************************************************************************************\
1074 * DescriptorMatcher *
1075 \****************************************************************************************/
1077 * Abstract base class for matching two sets of descriptors.
1079 class CV_EXPORTS_W DescriptorMatcher : public Algorithm
1082 virtual ~DescriptorMatcher();
1085 * Add descriptors to train descriptor collection.
1086 * descriptors Descriptors to add. Each descriptors[i] is a descriptors set from one image.
1088 CV_WRAP virtual void add( InputArrayOfArrays descriptors );
1090 * Get train descriptors collection.
1092 CV_WRAP const std::vector<Mat>& getTrainDescriptors() const;
1094 * Clear train descriptors collection.
1096 CV_WRAP virtual void clear();
1099 * Return true if there are not train descriptors in collection.
1101 CV_WRAP virtual bool empty() const;
1103 * Return true if the matcher supports mask in match methods.
1105 CV_WRAP virtual bool isMaskSupported() const = 0;
1108 * Train matcher (e.g. train flann index).
1109 * In all methods to match the method train() is run every time before matching.
1110 * Some descriptor matchers (e.g. BruteForceMatcher) have empty implementation
1111 * of this method, other matchers really train their inner structures
1112 * (e.g. FlannBasedMatcher trains flann::Index). So nonempty implementation
1113 * of train() should check the class object state and do traing/retraining
1114 * only if the state requires that (e.g. FlannBasedMatcher trains flann::Index
1115 * if it has not trained yet or if new descriptors have been added to the train
1118 CV_WRAP virtual void train();
1120 * Group of methods to match descriptors from image pair.
1121 * Method train() is run in this methods.
1123 // Find one best match for each query descriptor (if mask is empty).
1124 CV_WRAP void match( InputArray queryDescriptors, InputArray trainDescriptors,
1125 CV_OUT std::vector<DMatch>& matches, InputArray mask=noArray() ) const;
1126 // Find k best matches for each query descriptor (in increasing order of distances).
1127 // compactResult is used when mask is not empty. If compactResult is false matches
1128 // vector will have the same size as queryDescriptors rows. If compactResult is true
1129 // matches vector will not contain matches for fully masked out query descriptors.
1130 CV_WRAP void knnMatch( InputArray queryDescriptors, InputArray trainDescriptors,
1131 CV_OUT std::vector<std::vector<DMatch> >& matches, int k,
1132 InputArray mask=noArray(), bool compactResult=false ) const;
1133 // Find best matches for each query descriptor which have distance less than
1134 // maxDistance (in increasing order of distances).
1135 void radiusMatch( InputArray queryDescriptors, InputArray trainDescriptors,
1136 std::vector<std::vector<DMatch> >& matches, float maxDistance,
1137 InputArray mask=noArray(), bool compactResult=false ) const;
1139 * Group of methods to match descriptors from one image to image set.
1140 * See description of similar methods for matching image pair above.
1142 CV_WRAP void match( InputArray queryDescriptors, CV_OUT std::vector<DMatch>& matches,
1143 InputArrayOfArrays masks=noArray() );
1144 CV_WRAP void knnMatch( InputArray queryDescriptors, CV_OUT std::vector<std::vector<DMatch> >& matches, int k,
1145 InputArrayOfArrays masks=noArray(), bool compactResult=false );
1146 void radiusMatch( InputArray queryDescriptors, std::vector<std::vector<DMatch> >& matches, float maxDistance,
1147 InputArrayOfArrays masks=noArray(), bool compactResult=false );
1149 // Reads matcher object from a file node
1150 virtual void read( const FileNode& );
1151 // Writes matcher object to a file storage
1152 virtual void write( FileStorage& ) const;
1154 // Clone the matcher. If emptyTrainData is false the method create deep copy of the object, i.e. copies
1155 // both parameters and train data. If emptyTrainData is true the method create object copy with current parameters
1156 // but with empty train data.
1157 virtual Ptr<DescriptorMatcher> clone( bool emptyTrainData=false ) const = 0;
1159 CV_WRAP static Ptr<DescriptorMatcher> create( const String& descriptorMatcherType );
1162 * Class to work with descriptors from several images as with one merged matrix.
1163 * It is used e.g. in FlannBasedMatcher.
1165 class CV_EXPORTS DescriptorCollection
1168 DescriptorCollection();
1169 DescriptorCollection( const DescriptorCollection& collection );
1170 virtual ~DescriptorCollection();
1172 // Vector of matrices "descriptors" will be merged to one matrix "mergedDescriptors" here.
1173 void set( const std::vector<Mat>& descriptors );
1174 virtual void clear();
1176 const Mat& getDescriptors() const;
1177 const Mat getDescriptor( int imgIdx, int localDescIdx ) const;
1178 const Mat getDescriptor( int globalDescIdx ) const;
1179 void getLocalIdx( int globalDescIdx, int& imgIdx, int& localDescIdx ) const;
1184 Mat mergedDescriptors;
1185 std::vector<int> startIdxs;
1188 // In fact the matching is implemented only by the following two methods. These methods suppose
1189 // that the class object has been trained already. Public match methods call these methods
1190 // after calling train().
1191 virtual void knnMatchImpl( InputArray queryDescriptors, std::vector<std::vector<DMatch> >& matches, int k,
1192 InputArrayOfArrays masks=noArray(), bool compactResult=false ) = 0;
1193 virtual void radiusMatchImpl( InputArray queryDescriptors, std::vector<std::vector<DMatch> >& matches, float maxDistance,
1194 InputArrayOfArrays masks=noArray(), bool compactResult=false ) = 0;
1196 static bool isPossibleMatch( InputArray mask, int queryIdx, int trainIdx );
1197 static bool isMaskedOut( InputArrayOfArrays masks, int queryIdx );
1199 static Mat clone_op( Mat m ) { return m.clone(); }
1200 void checkMasks( InputArrayOfArrays masks, int queryDescriptorsCount ) const;
1202 // Collection of descriptors from train images.
1203 std::vector<Mat> trainDescCollection;
1204 std::vector<UMat> utrainDescCollection;
1208 * Brute-force descriptor matcher.
1210 * For each descriptor in the first set, this matcher finds the closest
1211 * descriptor in the second set by trying each one.
1213 * For efficiency, BruteForceMatcher is templated on the distance metric.
1214 * For float descriptors, a common choice would be cv::L2<float>.
1216 class CV_EXPORTS_W BFMatcher : public DescriptorMatcher
1219 CV_WRAP BFMatcher( int normType=NORM_L2, bool crossCheck=false );
1220 virtual ~BFMatcher() {}
1222 virtual bool isMaskSupported() const { return true; }
1224 virtual Ptr<DescriptorMatcher> clone( bool emptyTrainData=false ) const;
1226 AlgorithmInfo* info() const;
1228 virtual void knnMatchImpl( InputArray queryDescriptors, std::vector<std::vector<DMatch> >& matches, int k,
1229 InputArrayOfArrays masks=noArray(), bool compactResult=false );
1230 virtual void radiusMatchImpl( InputArray queryDescriptors, std::vector<std::vector<DMatch> >& matches, float maxDistance,
1231 InputArrayOfArrays masks=noArray(), bool compactResult=false );
1239 * Flann based matcher
1241 class CV_EXPORTS_W FlannBasedMatcher : public DescriptorMatcher
1244 CV_WRAP FlannBasedMatcher( const Ptr<flann::IndexParams>& indexParams=makePtr<flann::KDTreeIndexParams>(),
1245 const Ptr<flann::SearchParams>& searchParams=makePtr<flann::SearchParams>() );
1247 virtual void add( InputArrayOfArrays descriptors );
1248 virtual void clear();
1250 // Reads matcher object from a file node
1251 virtual void read( const FileNode& );
1252 // Writes matcher object to a file storage
1253 virtual void write( FileStorage& ) const;
1255 virtual void train();
1256 virtual bool isMaskSupported() const;
1258 virtual Ptr<DescriptorMatcher> clone( bool emptyTrainData=false ) const;
1260 AlgorithmInfo* info() const;
1262 static void convertToDMatches( const DescriptorCollection& descriptors,
1263 const Mat& indices, const Mat& distances,
1264 std::vector<std::vector<DMatch> >& matches );
1266 virtual void knnMatchImpl( InputArray queryDescriptors, std::vector<std::vector<DMatch> >& matches, int k,
1267 InputArrayOfArrays masks=noArray(), bool compactResult=false );
1268 virtual void radiusMatchImpl( InputArray queryDescriptors, std::vector<std::vector<DMatch> >& matches, float maxDistance,
1269 InputArrayOfArrays masks=noArray(), bool compactResult=false );
1271 Ptr<flann::IndexParams> indexParams;
1272 Ptr<flann::SearchParams> searchParams;
1273 Ptr<flann::Index> flannIndex;
1275 DescriptorCollection mergedDescriptors;
1279 /****************************************************************************************\
1280 * GenericDescriptorMatcher *
1281 \****************************************************************************************/
1283 * Abstract interface for a keypoint descriptor and matcher
1285 class GenericDescriptorMatcher;
1286 typedef GenericDescriptorMatcher GenericDescriptorMatch;
1288 class CV_EXPORTS GenericDescriptorMatcher
1291 GenericDescriptorMatcher();
1292 virtual ~GenericDescriptorMatcher();
1295 * Add train collection: images and keypoints from them.
1296 * images A set of train images.
1297 * ketpoints Keypoint collection that have been detected on train images.
1299 * Keypoints for which a descriptor cannot be computed are removed. Such keypoints
1300 * must be filtered in this method befor adding keypoints to train collection "trainPointCollection".
1301 * If inheritor class need perform such prefiltering the method add() must be overloaded.
1302 * In the other class methods programmer has access to the train keypoints by a constant link.
1304 virtual void add( InputArrayOfArrays images,
1305 std::vector<std::vector<KeyPoint> >& keypoints );
1307 const std::vector<Mat>& getTrainImages() const;
1308 const std::vector<std::vector<KeyPoint> >& getTrainKeypoints() const;
1311 * Clear images and keypoints storing in train collection.
1313 virtual void clear();
1315 * Returns true if matcher supports mask to match descriptors.
1317 virtual bool isMaskSupported() = 0;
1319 * Train some inner structures (e.g. flann index or decision trees).
1320 * train() methods is run every time in matching methods. So the method implementation
1321 * should has a check whether these inner structures need be trained/retrained or not.
1323 virtual void train();
1326 * Classifies query keypoints.
1327 * queryImage The query image
1328 * queryKeypoints Keypoints from the query image
1329 * trainImage The train image
1330 * trainKeypoints Keypoints from the train image
1332 // Classify keypoints from query image under one train image.
1333 void classify( InputArray queryImage, std::vector<KeyPoint>& queryKeypoints,
1334 InputArray trainImage, std::vector<KeyPoint>& trainKeypoints ) const;
1335 // Classify keypoints from query image under train image collection.
1336 void classify( InputArray queryImage, std::vector<KeyPoint>& queryKeypoints );
1339 * Group of methods to match keypoints from image pair.
1340 * Keypoints for which a descriptor cannot be computed are removed.
1341 * train() method is called here.
1343 // Find one best match for each query descriptor (if mask is empty).
1344 void match( InputArray queryImage, std::vector<KeyPoint>& queryKeypoints,
1345 InputArray trainImage, std::vector<KeyPoint>& trainKeypoints,
1346 std::vector<DMatch>& matches, InputArray mask=noArray() ) const;
1347 // Find k best matches for each query keypoint (in increasing order of distances).
1348 // compactResult is used when mask is not empty. If compactResult is false matches
1349 // vector will have the same size as queryDescriptors rows.
1350 // If compactResult is true matches vector will not contain matches for fully masked out query descriptors.
1351 void knnMatch( InputArray queryImage, std::vector<KeyPoint>& queryKeypoints,
1352 InputArray trainImage, std::vector<KeyPoint>& trainKeypoints,
1353 std::vector<std::vector<DMatch> >& matches, int k,
1354 InputArray mask=noArray(), bool compactResult=false ) const;
1355 // Find best matches for each query descriptor which have distance less than maxDistance (in increasing order of distances).
1356 void radiusMatch( InputArray queryImage, std::vector<KeyPoint>& queryKeypoints,
1357 InputArray trainImage, std::vector<KeyPoint>& trainKeypoints,
1358 std::vector<std::vector<DMatch> >& matches, float maxDistance,
1359 InputArray mask=noArray(), bool compactResult=false ) const;
1361 * Group of methods to match keypoints from one image to image set.
1362 * See description of similar methods for matching image pair above.
1364 void match( InputArray queryImage, std::vector<KeyPoint>& queryKeypoints,
1365 std::vector<DMatch>& matches, InputArrayOfArrays masks=noArray() );
1366 void knnMatch( InputArray queryImage, std::vector<KeyPoint>& queryKeypoints,
1367 std::vector<std::vector<DMatch> >& matches, int k,
1368 InputArrayOfArrays masks=noArray(), bool compactResult=false );
1369 void radiusMatch(InputArray queryImage, std::vector<KeyPoint>& queryKeypoints,
1370 std::vector<std::vector<DMatch> >& matches, float maxDistance,
1371 InputArrayOfArrays masks=noArray(), bool compactResult=false );
1373 // Reads matcher object from a file node
1374 virtual void read( const FileNode& fn );
1375 // Writes matcher object to a file storage
1376 virtual void write( FileStorage& fs ) const;
1378 // Return true if matching object is empty (e.g. feature detector or descriptor matcher are empty)
1379 virtual bool empty() const;
1381 // Clone the matcher. If emptyTrainData is false the method create deep copy of the object, i.e. copies
1382 // both parameters and train data. If emptyTrainData is true the method create object copy with current parameters
1383 // but with empty train data.
1384 virtual Ptr<GenericDescriptorMatcher> clone( bool emptyTrainData=false ) const = 0;
1386 static Ptr<GenericDescriptorMatcher> create( const String& genericDescritptorMatcherType,
1387 const String ¶msFilename=String() );
1390 // In fact the matching is implemented only by the following two methods. These methods suppose
1391 // that the class object has been trained already. Public match methods call these methods
1392 // after calling train().
1393 virtual void knnMatchImpl( InputArray queryImage, std::vector<KeyPoint>& queryKeypoints,
1394 std::vector<std::vector<DMatch> >& matches, int k,
1395 InputArrayOfArrays masks, bool compactResult ) = 0;
1396 virtual void radiusMatchImpl( InputArray queryImage, std::vector<KeyPoint>& queryKeypoints,
1397 std::vector<std::vector<DMatch> >& matches, float maxDistance,
1398 InputArrayOfArrays masks, bool compactResult ) = 0;
1400 * A storage for sets of keypoints together with corresponding images and class IDs
1402 class CV_EXPORTS KeyPointCollection
1405 KeyPointCollection();
1406 KeyPointCollection( const KeyPointCollection& collection );
1407 void add( const std::vector<Mat>& images, const std::vector<std::vector<KeyPoint> >& keypoints );
1410 // Returns the total number of keypoints in the collection
1411 size_t keypointCount() const;
1412 size_t imageCount() const;
1414 const std::vector<std::vector<KeyPoint> >& getKeypoints() const;
1415 const std::vector<KeyPoint>& getKeypoints( int imgIdx ) const;
1416 const KeyPoint& getKeyPoint( int imgIdx, int localPointIdx ) const;
1417 const KeyPoint& getKeyPoint( int globalPointIdx ) const;
1418 void getLocalIdx( int globalPointIdx, int& imgIdx, int& localPointIdx ) const;
1420 const std::vector<Mat>& getImages() const;
1421 const Mat& getImage( int imgIdx ) const;
1426 std::vector<Mat> images;
1427 std::vector<std::vector<KeyPoint> > keypoints;
1428 // global indices of the first points in each image, startIndices.size() = keypoints.size()
1429 std::vector<int> startIndices;
1432 static Mat clone_op( Mat m ) { return m.clone(); }
1435 KeyPointCollection trainPointCollection;
1439 /****************************************************************************************\
1440 * VectorDescriptorMatcher *
1441 \****************************************************************************************/
1444 * A class used for matching descriptors that can be described as vectors in a finite-dimensional space
1446 class VectorDescriptorMatcher;
1447 typedef VectorDescriptorMatcher VectorDescriptorMatch;
1449 class CV_EXPORTS VectorDescriptorMatcher : public GenericDescriptorMatcher
1452 VectorDescriptorMatcher( const Ptr<DescriptorExtractor>& extractor, const Ptr<DescriptorMatcher>& matcher );
1453 virtual ~VectorDescriptorMatcher();
1455 virtual void add( InputArrayOfArrays imgCollection,
1456 std::vector<std::vector<KeyPoint> >& pointCollection );
1458 virtual void clear();
1460 virtual void train();
1462 virtual bool isMaskSupported();
1464 virtual void read( const FileNode& fn );
1465 virtual void write( FileStorage& fs ) const;
1466 virtual bool empty() const;
1468 virtual Ptr<GenericDescriptorMatcher> clone( bool emptyTrainData=false ) const;
1471 virtual void knnMatchImpl( InputArray queryImage, std::vector<KeyPoint>& queryKeypoints,
1472 std::vector<std::vector<DMatch> >& matches, int k,
1473 InputArrayOfArrays masks, bool compactResult );
1474 virtual void radiusMatchImpl( InputArray queryImage, std::vector<KeyPoint>& queryKeypoints,
1475 std::vector<std::vector<DMatch> >& matches, float maxDistance,
1476 InputArrayOfArrays masks, bool compactResult );
1478 Ptr<DescriptorExtractor> extractor;
1479 Ptr<DescriptorMatcher> matcher;
1482 /****************************************************************************************\
1483 * Drawing functions *
1484 \****************************************************************************************/
1485 struct CV_EXPORTS DrawMatchesFlags
1487 enum{ DEFAULT = 0, // Output image matrix will be created (Mat::create),
1488 // i.e. existing memory of output image may be reused.
1489 // Two source image, matches and single keypoints will be drawn.
1490 // For each keypoint only the center point will be drawn (without
1491 // the circle around keypoint with keypoint size and orientation).
1492 DRAW_OVER_OUTIMG = 1, // Output image matrix will not be created (Mat::create).
1493 // Matches will be drawn on existing content of output image.
1494 NOT_DRAW_SINGLE_POINTS = 2, // Single keypoints will not be drawn.
1495 DRAW_RICH_KEYPOINTS = 4 // For each keypoint the circle around keypoint with keypoint size and
1496 // orientation will be drawn.
1501 CV_EXPORTS_W void drawKeypoints( InputArray image, const std::vector<KeyPoint>& keypoints, InputOutputArray outImage,
1502 const Scalar& color=Scalar::all(-1), int flags=DrawMatchesFlags::DEFAULT );
1504 // Draws matches of keypints from two images on output image.
1505 CV_EXPORTS_W void drawMatches( InputArray img1, const std::vector<KeyPoint>& keypoints1,
1506 InputArray img2, const std::vector<KeyPoint>& keypoints2,
1507 const std::vector<DMatch>& matches1to2, InputOutputArray outImg,
1508 const Scalar& matchColor=Scalar::all(-1), const Scalar& singlePointColor=Scalar::all(-1),
1509 const std::vector<char>& matchesMask=std::vector<char>(), int flags=DrawMatchesFlags::DEFAULT );
1511 CV_EXPORTS_AS(drawMatchesKnn) void drawMatches( InputArray img1, const std::vector<KeyPoint>& keypoints1,
1512 InputArray img2, const std::vector<KeyPoint>& keypoints2,
1513 const std::vector<std::vector<DMatch> >& matches1to2, InputOutputArray outImg,
1514 const Scalar& matchColor=Scalar::all(-1), const Scalar& singlePointColor=Scalar::all(-1),
1515 const std::vector<std::vector<char> >& matchesMask=std::vector<std::vector<char> >(), int flags=DrawMatchesFlags::DEFAULT );
1517 /****************************************************************************************\
1518 * Functions to evaluate the feature detectors and [generic] descriptor extractors *
1519 \****************************************************************************************/
1521 CV_EXPORTS void evaluateFeatureDetector( const Mat& img1, const Mat& img2, const Mat& H1to2,
1522 std::vector<KeyPoint>* keypoints1, std::vector<KeyPoint>* keypoints2,
1523 float& repeatability, int& correspCount,
1524 const Ptr<FeatureDetector>& fdetector=Ptr<FeatureDetector>() );
1526 CV_EXPORTS void computeRecallPrecisionCurve( const std::vector<std::vector<DMatch> >& matches1to2,
1527 const std::vector<std::vector<uchar> >& correctMatches1to2Mask,
1528 std::vector<Point2f>& recallPrecisionCurve );
1530 CV_EXPORTS float getRecall( const std::vector<Point2f>& recallPrecisionCurve, float l_precision );
1531 CV_EXPORTS int getNearestPoint( const std::vector<Point2f>& recallPrecisionCurve, float l_precision );
1533 CV_EXPORTS void evaluateGenericDescriptorMatcher( const Mat& img1, const Mat& img2, const Mat& H1to2,
1534 std::vector<KeyPoint>& keypoints1, std::vector<KeyPoint>& keypoints2,
1535 std::vector<std::vector<DMatch> >* matches1to2, std::vector<std::vector<uchar> >* correctMatches1to2Mask,
1536 std::vector<Point2f>& recallPrecisionCurve,
1537 const Ptr<GenericDescriptorMatcher>& dmatch=Ptr<GenericDescriptorMatcher>() );
1540 /****************************************************************************************\
1541 * Bag of visual words *
1542 \****************************************************************************************/
1544 * Abstract base class for training of a 'bag of visual words' vocabulary from a set of descriptors
1546 class CV_EXPORTS BOWTrainer
1550 virtual ~BOWTrainer();
1552 void add( const Mat& descriptors );
1553 const std::vector<Mat>& getDescriptors() const;
1554 int descriptorsCount() const;
1556 virtual void clear();
1559 * Train visual words vocabulary, that is cluster training descriptors and
1560 * compute cluster centers.
1561 * Returns cluster centers.
1563 * descriptors Training descriptors computed on images keypoints.
1565 virtual Mat cluster() const = 0;
1566 virtual Mat cluster( const Mat& descriptors ) const = 0;
1569 std::vector<Mat> descriptors;
1574 * This is BOWTrainer using cv::kmeans to get vocabulary.
1576 class CV_EXPORTS BOWKMeansTrainer : public BOWTrainer
1579 BOWKMeansTrainer( int clusterCount, const TermCriteria& termcrit=TermCriteria(),
1580 int attempts=3, int flags=KMEANS_PP_CENTERS );
1581 virtual ~BOWKMeansTrainer();
1583 // Returns trained vocabulary (i.e. cluster centers).
1584 virtual Mat cluster() const;
1585 virtual Mat cluster( const Mat& descriptors ) const;
1590 TermCriteria termcrit;
1596 * Class to compute image descriptor using bag of visual words.
1598 class CV_EXPORTS BOWImgDescriptorExtractor
1601 BOWImgDescriptorExtractor( const Ptr<DescriptorExtractor>& dextractor,
1602 const Ptr<DescriptorMatcher>& dmatcher );
1603 BOWImgDescriptorExtractor( const Ptr<DescriptorMatcher>& dmatcher );
1604 virtual ~BOWImgDescriptorExtractor();
1606 void setVocabulary( const Mat& vocabulary );
1607 const Mat& getVocabulary() const;
1608 void compute( InputArray image, std::vector<KeyPoint>& keypoints, OutputArray imgDescriptor,
1609 std::vector<std::vector<int> >* pointIdxsOfClusters=0, Mat* descriptors=0 );
1610 void compute( InputArray keypointDescriptors, OutputArray imgDescriptor,
1611 std::vector<std::vector<int> >* pointIdxsOfClusters=0 );
1612 // compute() is not constant because DescriptorMatcher::match is not constant
1614 int descriptorSize() const;
1615 int descriptorType() const;
1619 Ptr<DescriptorExtractor> dextractor;
1620 Ptr<DescriptorMatcher> dmatcher;
1623 } /* namespace cv */