opencv/modules/features2d/perf/perf_feature2d.hpp

88 lines
3.6 KiB
C++
Raw Normal View History

Merge pull request #8869 from hrnr:akaze_part1 [GSOC] Speeding-up AKAZE, part #1 (#8869) * ts: expand arguments before stringifications in CV_ENUM and CV_FLAGS added protective macros to always force macro expansion of arguments. This allows using CV_ENUM and CV_FLAGS with macro arguments. * feature2d: unify perf test use the same test for all detectors/descriptors we have. * added AKAZE tests * features2d: extend perf tests * add BRISK, KAZE, MSER * run all extract tests on AKAZE keypoints, so that the test si more comparable for the speed of extraction * feature2d: rework opencl perf tests use the same configuration as cpu tests * feature2d: fix descriptors allocation for AKAZE and KAZE fix crash when descriptors are UMat * feature2d: name enum to fix build with older gcc * Revert "ts: expand arguments before stringifications in CV_ENUM and CV_FLAGS" This reverts commit 19538cac1e45b0cec98190cf06a5ecb07d9b596e. This wasn't a great idea after all. There is a lot of flags implemented as #define, that we don't want to expand. * feature2d: fix expansion problems with CV_ENUM in perf * expand arguments before passing them to CV_ENUM. This does not need modifications of CV_ENUM. * added include guards to `perf_feature2d.hpp` * feature2d: fix crash in AKAZE when using KAZE descriptors * out-of-bound access in Get_MSURF_Descriptor_64 * this happened reliably when running on provided keypoints (not computed by the same instance) * feature2d: added regression tests for AKAZE * test with both MLDB and KAZE keypoints * feature2d: do not compute keypoints orientation twice * always compute keypoints orientation, when computing keypoints * do not recompute keypoint orientation when computing descriptors this allows to test detection and extraction separately * features2d: fix crash in AKAZE * out-of-bound reads near the image edge * same as the bug in KAZE descriptors * feature2d: refactor invariance testing * split detectors and descriptors tests * rewrite to google test to simplify debugging * add tests for AKAZE and one test for ORB * stitching: add tests with AKAZE feature finder * added basic stitching cpu and ocl tests * fix bug in AKAZE wrapper for stitching pipeline causing lots of ! OPENCV warning: getUMat()/getMat() call chain possible problem. ! Base object is dead, while nested/derived object is still alive or processed. ! Please check lifetime of UMat/Mat objects!
2017-06-21 19:33:09 +08:00
#ifndef __OPENCV_PERF_FEATURE2D_HPP__
#define __OPENCV_PERF_FEATURE2D_HPP__
#include "perf_precomp.hpp"
/* cofiguration for tests of detectors/descriptors. shared between ocl and cpu tests. */
using namespace std;
using namespace cv;
using namespace perf;
using std::tr1::make_tuple;
using std::tr1::get;
// detectors/descriptors configurations to test
#define DETECTORS_ONLY \
FAST_DEFAULT, FAST_20_TRUE_TYPE5_8, FAST_20_TRUE_TYPE7_12, FAST_20_TRUE_TYPE9_16, \
FAST_20_FALSE_TYPE5_8, FAST_20_FALSE_TYPE7_12, FAST_20_FALSE_TYPE9_16, \
\
AGAST_DEFAULT, AGAST_5_8, AGAST_7_12d, AGAST_7_12s, AGAST_OAST_9_16, \
\
MSER_DEFAULT
#define DETECTORS_EXTRACTORS \
ORB_DEFAULT, ORB_1500_13_1, \
AKAZE_DEFAULT, AKAZE_DESCRIPTOR_KAZE, \
BRISK_DEFAULT, \
KAZE_DEFAULT
#define CV_ENUM_EXPAND(name, ...) CV_ENUM(name, __VA_ARGS__)
enum Feature2DVals { DETECTORS_ONLY, DETECTORS_EXTRACTORS };
CV_ENUM_EXPAND(Feature2DType, DETECTORS_ONLY, DETECTORS_EXTRACTORS)
typedef std::tr1::tuple<Feature2DType, string> Feature2DType_String_t;
typedef perf::TestBaseWithParam<Feature2DType_String_t> feature2d;
#define TEST_IMAGES testing::Values(\
"cv/detectors_descriptors_evaluation/images_datasets/leuven/img1.png",\
Merge pull request #8951 from hrnr:akaze_part2 [GSOC] Speeding-up AKAZE, part #2 (#8951) * feature2d: instrument more functions used in AKAZE * rework Compute_Determinant_Hessian_Response * this takes 84% of time of Feature_Detection * run everything in parallel * compute Scharr kernels just once * compute sigma more efficiently * allocate all matrices in evolution without zeroing * features2d: add one bigger image to tests * now test have images: 600x768, 900x600 and 1385x700 to cover different resolutions * explicitly zero Lx and Ly * add Lflow and Lstep to evolution as in original AKAZE code * reworked computing keypoints orientation integrated faster function from https://github.com/h2suzuki/fast_akaze * use standard fastAtan2 instead of getAngle * compute keypoints orientation in parallel * fix visual studio warnings * replace some wrapped functions with direct calls to OpenCV functions * improved readability for people familiar with opencv * do not same image twice in base level * rework diffusity stencil * use one pass stencil for diffusity from https://github.com/h2suzuki/fast_akaze * improve locality in Create_Scale_Space * always compute determinat od hessian and spacial derivatives * this needs to be computed always as we need derivatives while computing descriptors * fixed tests of AKAZE with KAZE descriptors which have been affected by this Currently it computes all first and second order derivatives together and the determiant of the hessian. For descriptors it would be enough to compute just first order derivates, but it is not probably worth it optimize for scenario where descriptors and keypoints are computed separately, since it is already very inefficient. When computing keypoint and descriptors together it is faster to do it the current way (preserves locality). * parallelize non linear diffusion computation * do multiplication right in the nlp diffusity kernel * rework kfactor computation * get rid of sharing buffers when creating scale space pyramid, the performace impact is neglegible * features2d: initialize TBB scheduler in perf tests * ensures more stable output * more reasonable profiles, since the first call of parallel_for_ is not getting big performace hit * compute_kfactor: interleave finding of maximum and computing distance * no need to go twice through the data * start to use UMats in AKAZE to leverage OpenCl in the future * fixed bug that prevented computing determinant for scale pyramid of size 1 (just the base image) * all descriptors now support writing to uninitialized memory * use InputArray and OutputArray for input image and descriptors, allows to make use UMAt that user passes to us * enable use of all existing ocl paths in AKAZE * all parts that uses ocl-enabled functions should use ocl by now * imgproc: fix dispatching of IPP version when OCL is disabled * when OCL is disabled IPP version should be always prefered (even when the dst is UMat) * get rid of copy in DeterminantHessian response * this slows CPU version considerably * do no run in parallel when running with OCL * store derivations as UMat in pyramid * enables OCL path computing of determint hessian * will allow to compute descriptors on GPU in the future * port diffusivity to OCL * diffusivity itself is not a blocker, but this saves us downloading and uploading derivations * implement kernel for nonlinear scalar diffusion step * download the pyramid from GPU just once we don't want to downlaod matrices ad hoc from gpu when the function in AKAZE needs it. There is a HUGE mapping overhead and without shared memory support a LOT of unnecessary transfers. This maps/downloads matrices just once. * fix bug with uninitialized values in non linear diffusion * this was causing spurious segfaults in stitching tests due to propagation of NaNs * added new test, which checks for NaNs (added new debug asserts for NaNs) * valgrind now says everything is ok * add nonlinear diffusion step OCL implementation * Lt in pyramid changed to UMat, it will be downlaoded from GPU along with Lx, Ly * fix bug in pm_g2 kernel. OpenCV mangles dimensions passed to OpenCL, so we need to check for boundaries in each OCL kernel. * port computing of determinant to OCL * computing of determinant is not a blocker, but with this change we don't need to download all spatial derivatives to CPU, we only download determinant * make Ldet in the pyramid UMat, download it from CPU together with the other parts of the pyramid * add profiling macros * fix visual studio warning * instrument non_linear_diffusion * remove changes I have made to TEvolution * TEvolution is used only in KAZE now * Revert "features2d: initialize TBB scheduler in perf tests" This reverts commit ba81e2a711ae009ce3c5459775627b6423112669.
2017-08-01 20:46:01 +08:00
"stitching/a3.png", \
"stitching/s2.jpg")
Merge pull request #8869 from hrnr:akaze_part1 [GSOC] Speeding-up AKAZE, part #1 (#8869) * ts: expand arguments before stringifications in CV_ENUM and CV_FLAGS added protective macros to always force macro expansion of arguments. This allows using CV_ENUM and CV_FLAGS with macro arguments. * feature2d: unify perf test use the same test for all detectors/descriptors we have. * added AKAZE tests * features2d: extend perf tests * add BRISK, KAZE, MSER * run all extract tests on AKAZE keypoints, so that the test si more comparable for the speed of extraction * feature2d: rework opencl perf tests use the same configuration as cpu tests * feature2d: fix descriptors allocation for AKAZE and KAZE fix crash when descriptors are UMat * feature2d: name enum to fix build with older gcc * Revert "ts: expand arguments before stringifications in CV_ENUM and CV_FLAGS" This reverts commit 19538cac1e45b0cec98190cf06a5ecb07d9b596e. This wasn't a great idea after all. There is a lot of flags implemented as #define, that we don't want to expand. * feature2d: fix expansion problems with CV_ENUM in perf * expand arguments before passing them to CV_ENUM. This does not need modifications of CV_ENUM. * added include guards to `perf_feature2d.hpp` * feature2d: fix crash in AKAZE when using KAZE descriptors * out-of-bound access in Get_MSURF_Descriptor_64 * this happened reliably when running on provided keypoints (not computed by the same instance) * feature2d: added regression tests for AKAZE * test with both MLDB and KAZE keypoints * feature2d: do not compute keypoints orientation twice * always compute keypoints orientation, when computing keypoints * do not recompute keypoint orientation when computing descriptors this allows to test detection and extraction separately * features2d: fix crash in AKAZE * out-of-bound reads near the image edge * same as the bug in KAZE descriptors * feature2d: refactor invariance testing * split detectors and descriptors tests * rewrite to google test to simplify debugging * add tests for AKAZE and one test for ORB * stitching: add tests with AKAZE feature finder * added basic stitching cpu and ocl tests * fix bug in AKAZE wrapper for stitching pipeline causing lots of ! OPENCV warning: getUMat()/getMat() call chain possible problem. ! Base object is dead, while nested/derived object is still alive or processed. ! Please check lifetime of UMat/Mat objects!
2017-06-21 19:33:09 +08:00
static inline Ptr<Feature2D> getFeature2D(Feature2DType type)
{
switch(type) {
case ORB_DEFAULT:
return ORB::create();
case ORB_1500_13_1:
return ORB::create(1500, 1.3f, 1);
case FAST_DEFAULT:
return FastFeatureDetector::create();
case FAST_20_TRUE_TYPE5_8:
return FastFeatureDetector::create(20, true, FastFeatureDetector::TYPE_5_8);
case FAST_20_TRUE_TYPE7_12:
return FastFeatureDetector::create(20, true, FastFeatureDetector::TYPE_7_12);
case FAST_20_TRUE_TYPE9_16:
return FastFeatureDetector::create(20, true, FastFeatureDetector::TYPE_9_16);
case FAST_20_FALSE_TYPE5_8:
return FastFeatureDetector::create(20, false, FastFeatureDetector::TYPE_5_8);
case FAST_20_FALSE_TYPE7_12:
return FastFeatureDetector::create(20, false, FastFeatureDetector::TYPE_7_12);
case FAST_20_FALSE_TYPE9_16:
return FastFeatureDetector::create(20, false, FastFeatureDetector::TYPE_9_16);
case AGAST_DEFAULT:
return AgastFeatureDetector::create();
case AGAST_5_8:
return AgastFeatureDetector::create(70, true, AgastFeatureDetector::AGAST_5_8);
case AGAST_7_12d:
return AgastFeatureDetector::create(70, true, AgastFeatureDetector::AGAST_7_12d);
case AGAST_7_12s:
return AgastFeatureDetector::create(70, true, AgastFeatureDetector::AGAST_7_12s);
case AGAST_OAST_9_16:
return AgastFeatureDetector::create(70, true, AgastFeatureDetector::OAST_9_16);
case AKAZE_DEFAULT:
return AKAZE::create();
case AKAZE_DESCRIPTOR_KAZE:
return AKAZE::create(AKAZE::DESCRIPTOR_KAZE);
case BRISK_DEFAULT:
return BRISK::create();
case KAZE_DEFAULT:
return KAZE::create();
case MSER_DEFAULT:
return MSER::create();
default:
return Ptr<Feature2D>();
}
}
#endif // __OPENCV_PERF_FEATURE2D_HPP__