opencv

Commit Graph

Author	SHA1	Message	Date
Alexander Alekhin	a4a47b538c	build: detect Android via '__ANDROID__' macro https://sourceforge.net/p/predef/wiki/OperatingSystems	8 years ago
Maksim Shabunin	a769d69a9d	Fixed several issues found by static analysis	8 years ago
Maksim Shabunin	32d4af36e2	Fixing some static analysis issues	8 years ago
LukeZhu	cc47ee3b43	Modify the pyrlk.cl to support winSize from 88 to 2424 for optical flow	8 years ago
LukeZhu	65be9e1978	Fix OpenCL's wrong output of calcOpticalFlowPyrLK function's output vector of err. Improve err's precison of the calcOpticalFlowPyrLK OpenCL function and add the relative test.	8 years ago
Michele Cancilla	9405c6d206	Improvement of array of equivalences’ upper bound + fix some wrong comments	8 years ago
Fernando Iglesias Garcia	21e698053f	Initialize background model in background subtraction KNN. The method does cvCheckPixelBackgroundNP (which reads bgmodel) before it is ever updated (in cvUpdatePixelBackgroundNP). This initialization is thus needed to avoid reads of unitialized values.	8 years ago
Utkarsh Sinha	b9c2d88e69	If the user passes an empty output matrix, initialize it to identity.	8 years ago
Vitaly Tuzov	bf5b7843e8	Extended set of OpenVX HAL calls disabled for small images	8 years ago
Matthias Grundmann	13540bf7f4	Update bgfg_gaussmix2.cpp Addressed comments.	8 years ago
Matthias Grundmann	7295dd7dec	Update bgfg_gaussmix2.cpp 1e-12 -> DBL_EPSILON	8 years ago
Matthias Grundmann	0a1767a6b5	Update bgfg_gaussmix2.cpp to avoid divide by zero cases.	8 years ago
Vitaly Tuzov	9a4b5a4545	OpenVX calls updated to use single common OpenVX context per thread	8 years ago
Rostislav Vasilikhin	328151c8a4	don't run OpenVX pyrLK if user asked for err vector	8 years ago
Rostislav Vasilikhin	5372c8b331	Merge pull request #7774 from savuor:openvx_pyrlk OpenVX optical flow PyrLK wrappers added (#7774) * wrappers for vx_pyramid added * initial version of Optical Flow PyrLK wrappers added * array downloading code simplified * disabled due to bad accuracy; fixed bugs, e.g. vendor-specific ones * rewritten for new macro use	8 years ago
Kotrix	10ccd0328c	Removed redundant calculations Operations inside for loops are unnecessary. They are repeated anyway just below for loops	8 years ago
mshabunin	3e28d51779	Fixed several OpenCL compiler warnings	8 years ago
Tomoaki Teshima	ea4ff91861	fix error from performance test of LK pyramid	8 years ago
Tomoaki Teshima	1ef740fa2c	use universal intrinsic implementation for calcSharrDeriv	8 years ago
Sean McBride	5357e28a2f	Removed some extra semi-colons Fixes clang -Wextra-semi warnings.	8 years ago
Pavel Rojtberg	27ae1578ac	BackgroundSubtractorMOG2Impl: only use OpenCL if input image is UMat	8 years ago
Pavel Vlasov	30a6cee2fe	Instrumentation for OpenCV API regions and IPP functions;	8 years ago
Vitaly Tuzov	d8dc6caf09	Fix for incorrect calcOpticalFlowPyrLK result evaluation with ARM NEON	8 years ago
k-shinotsuka	042b0a584d	improve LKTrackerInvoker::operator()	8 years ago
Gilles Rochefort	9e53f03968	Fixed BackgroundSubstractorMOG2 in opencv_video. The number of gaussians involved in a mixture is supposed to be dynamically adjusted. After being increased, the number of gaussians can't be reduced anymore. It seems to be a regression as the legacy code located in modules/legacy/src/bgfg_gaussmix.cpp allows to reduce such number of gaussians.	9 years ago
Gilles Rochefort	78de75397e	Fixed BackgroundSubstractorMOG2 in opencv_video. The number of gaussians involved in a mixture is supposed to be dynamically adjusted. After being increased, the number of gaussians can't be reduced anymore. It seems to be a regression as the legacy code located in modules/legacy/src/bgfg_gaussmix.cpp allows to reduce such number of gaussians.	9 years ago
mvukad	695e33b25b	Fix missing format when writing Algorithm-based objects Added a writeFormat() method to Algorithm which must be called by the write() method of derived classes.	9 years ago
Matthieu FT	78475a47a5	fix: bg substraction for float images with OpenCL	9 years ago
alcinos	e22b838af8	Wrap SparseOptFlow class around PyrLK optical flow computation	9 years ago
alcinos	6e3b90de9b	Add static creator for TVL1 optical flow class	9 years ago
alcinos	be4312ec3d	Wrap DenseOptFlow class around Farneback optical flow computation	9 years ago
Maksim Shabunin	832a03821d	Valgrind: do not use uninitialized data in optflow	9 years ago
Maksim Shabunin	6e9d0d9a0c	Visual Studio 2015 warning and test fixes	9 years ago
Alexander Alekhin	7213e5f68a	ocl: correct disabling of OpenCL code	9 years ago
Alexander Alekhin	ad70ab404c	ocl: workaround for getUMat()	9 years ago
StevenPuttemans	dd8589c352	change as suggested by original programmer and user of ecc software	9 years ago
Yan Wang	a00eb1e9f5	Move coefficient (32.0f) to the end of every iteration. It could reduce 7 multiplications every iteration. Signed-off-by: Yan Wang <yan.wang@linux.intel.com>	10 years ago
Yan Wang	2c1650ad33	Optimize pyrlk. 1. Remove uncessary index calculation. 2. Use mad/mad24 as possible. Signed-off-by: Yan Wang <yan.wang@linux.intel.com>	10 years ago
Vadim Pisarevsky	a4073ed676	made improvements in ECC code (comments from https://github.com/Itseez/opencv/pull/3845 ), as well as minor tweak in pthread's based parallel for.	10 years ago
Owen Healy	86fb9f8409	Modify findTransformECC to support a mask of pixels to consider Tests of the mask are also included. This is useful for registering a non-square image against a non-square template. This also needs to relax a sanity check as per https://github.com/Itseez/opencv/pull/3851	10 years ago
Owen Healy	ff48387a8a	Fix bug of uninitialized matrix in findTransformECC The matrix templateZM needs to be initialized because otherwise uninitialized values leak into the correlation in: const double correlation = templateZM.dot(imageWarped) In the worst case this will lead the correlation to be NaN ruining the whole routine. The subtraction does not initialize templateZM due to the mask. Unfortunately, the uninitialized values (by altering the correlation) have the side effect of dragging out the computation a little longer giving a slightly better error bound. This means that fixing this bug breaks perf_ecc where SANITY_CHECK(warpMat, 1e-3); is just a little too tight and happens to work due to the uninitialized values. Since this is a performance not a accuracy test I think it is OK to just relax the error bound a little bit (the tight error bound being after all the result of a bug).	10 years ago
Maksim Shabunin	ee2d7a1f39	Fixed warnings for iOS	10 years ago
Jonathan Viney	4f24dc0959	Fix KNN background subtraction not honoring width step.	10 years ago
Vladislav Vinogradov	cda6fed41f	move tegra namespace out of cv to prevent conflicts	10 years ago
Vladislav Vinogradov	44e41baffe	use new functions before all tegra:: calls	10 years ago
Maksim Shabunin	da383e65e2	Remove deprecated methods from cv::Algorithm	10 years ago
Firat Kalaycilar	2bda7aaf61	minor fix to make "updateBackgroundModel=false" option meaningful.	10 years ago
Tomasz Posłuszny	8dd213c829	BackgroundSubtractorMOG2, learningRate=0 optimization	10 years ago
Vadim Pisarevsky	d2b9dc5530	quickly corrected the previous refactoring of features2d: moved from set(SOME_PROP, val) to setSomeProp(val)	10 years ago
Pavel Vlasov	45958eaabc	Implementation detector and selector for IPP and OpenCL; IPP can be switched on and off on runtime; Optional implementation collector was added (switched off by default in CMake). Gathers data of implementation used in functions and report this info through performance TS; TS modifications for implementations control;	10 years ago

1 2 3 4 5 ...

343 Commits (9c8626bf3cc74ec42d7d0583c484eef444a338a0)