Commit Graph

11813 Commits

Author SHA1 Message Date
Vadim Pisarevsky
97bdc92d1b Merge pull request #3870 from Dmitry-Me:reduceVariableScope9 2015-03-25 13:32:33 +00:00
Vadim Pisarevsky
0545aeb11b Merge pull request #3670 from obilaniu:master 2015-03-25 10:10:05 +00:00
Vadim Pisarevsky
bd7c78ffd5 Merge pull request #3866 from tdsmith:python-linkage 2015-03-25 09:51:20 +00:00
Dmitry-Me
6318409a5d Reduce variable scope 2015-03-25 10:35:11 +03:00
Vadim Pisarevsky
5501cfd809 Merge pull request #3814 from erikrk:denoising-16bit-master 2015-03-24 15:23:58 +00:00
Vadim Pisarevsky
3c51fdffac Merge pull request #3849 from Dikay900:fix_tilted_features_crash 2015-03-24 15:19:13 +00:00
Vadim Pisarevsky
f7f10a7d2c Merge pull request #3854 from Dmitry-Me:reduceVariableScope8 2015-03-24 15:13:34 +00:00
Vadim Pisarevsky
d7cb99254e Merge pull request #3858 from MSOpenTech:fix-ml-test 2015-03-24 15:09:53 +00:00
Vadim Pisarevsky
70ff448067 Merge pull request #3859 from NewProggie:fix-doc-typo 2015-03-24 15:08:03 +00:00
Vadim Pisarevsky
54deb5ed5a Merge pull request #3860 from theodr:document_hconcat_vconcat 2015-03-24 15:07:34 +00:00
theodore
02c3d8d792 documentation for the hconcat and vconcat functions 2015-03-24 13:59:09 +01:00
Tim D. Smith
b9b743bbe3 Don't explicitly link Python on OS X
Explicitly linking to a Python framework on OS X prevents modules from
being built against one Python (i.e. system python) and imported from
another (i.e. Homebrew python); the interpreter segfaults if there's a
linkage to a foreign Python. Building the module with `-undefined
dynamic_lookup` instead of an explicit link allows the symbols to be
resolved at load time from a compatible python.
2015-03-23 23:36:46 -07:00
Erik Karlsson
01d3df0d00 Added normType parameter to fastNlMeansDenoisingMulti 2015-03-24 02:01:31 +01:00
Erik Karlsson
70a64ebe72 Added test cases 2015-03-24 01:16:17 +01:00
Erik Karlsson
5f8d688664 Added parameter normType to fastNlMeansDenoising 2015-03-24 00:47:11 +01:00
Erik Karlsson
9fff7896c5 Removed fastNlMeansDenoisingColored[Multi]Abs 2015-03-23 23:23:35 +01:00
Yan Wang
433c3da7cf Optimize the performance of cascade OpenCL kernel.
1. Use built-in mad() instead of += and *.
2. For stump stages, if weight.z == 0, avoid unnecessary calculation
because some features only have 2 rectangles..

It could improve
OCL_Cascade_Image_MinSize_CascadeClassifier.CascadeClassifier/*
about 10% mean values.

Signed-off-by: Yan Wang <yan.wang@linux.intel.com>
2015-03-23 17:47:40 +08:00
Vadim Pisarevsky
5a0c7a1330 Merge pull request #3853 from alalek:ocl_fix_bufferpool_cleanup 2015-03-23 08:10:22 +00:00
Vadim Pisarevsky
7ab7322915 Merge pull request #3848 from yxiong:fix-4125 2015-03-23 08:04:41 +00:00
Vadim Pisarevsky
daa99514fa Merge pull request #3846 from mshabunin:fix-headers 2015-03-23 08:01:25 +00:00
Kai Wolf
f815bbab24 Fix typo in types_c.h 2015-03-22 10:55:54 +01:00
Evgeny Agafonchikov
91bd7df8c1 Fixing invalid opencv_test_ml calls 2015-03-20 17:49:03 +03:00
Dikay900
fb8478eb7b add resetting the offset variable after setting a new image 2015-03-20 13:26:13 +01:00
Dmitry-Me
ce167e233b Reduce variable scope 2015-03-20 13:27:08 +03:00
Alexander Alekhin
b5529aef9c ocl/bufferpool: fix cleanup 2015-03-20 13:21:10 +03:00
Vadim Pisarevsky
8616cd875b Merge pull request #3836 from Dmitry-Me:uninitializedMember 2015-03-20 10:03:11 +00:00
Vadim Pisarevsky
072a4b5497 Merge pull request #3851 from ellbur:findTransformECC-initialize-bug 2015-03-20 09:34:04 +00:00
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).
2015-03-19 20:50:49 -04:00
Dikay900
80069f384a resolve crash of haarclassifier when using a tilted feature 2015-03-19 20:47:33 +01:00
Ying Xiong
decbdd1abb Fix compiler error on std::swap (CUDA7.0/clang/OSX) 2015-03-19 14:23:44 -04:00
Vadim Pisarevsky
9683f93a76 Merge pull request #3840 from MSOpenTech:concurrency-contrib 2015-03-19 15:11:01 +00:00
Vadim Pisarevsky
9997e6d337 Merge pull request #3823 from lupustr3:pvlasov/implementation_detector_update 2015-03-19 12:47:22 +00:00
Maksim Shabunin
9f08310307 Disabled TrainData::getNormCatValues wrapping due to pointer in parameters 2015-03-19 15:45:34 +03:00
Maksim Shabunin
5709ec85d7 Disabled missing python method for static ml function 2015-03-19 15:11:49 +03:00
Maksim Shabunin
5247eacc3c Add python and java support for ml module 2015-03-19 13:22:30 +03:00
Maksim Shabunin
d5b954c2d4 Moved property implementation macros to private header 2015-03-18 17:33:12 +03:00
Maksim Shabunin
7335a40a61 Replaced CV_PURE_PROPERTY macros with corresponding code 2015-03-18 17:23:42 +03:00
Maxim Kostin
d1a8171011 Added WinRT concurrency support. 2015-03-18 15:13:08 +03:00
Olexa Bilaniuk
8a1d8c8470 Merge remote-tracking branch 'refs/remotes/upstream/master' 2015-03-18 02:21:36 -04:00
Olexa Bilaniuk
6d27d488bf Bugfix in n* optimization.
Similar to the problem in LevMarq, arg.inl was being used instead of
best.inl. This opened us up to a potential segfault.
2015-03-18 02:21:16 -04:00
Philipp Hasper
e0efda7145 Fixed persistent.hpp documentation
Small error - object fs does not exist in this code example
2015-03-17 15:28:37 +01:00
Dmitry-Me
0c5faa6d24 Fix uninitialized members, fix log output typo 2015-03-17 14:47:00 +03:00
Vadim Pisarevsky
903f10c237 Merge pull request #3804 from polarfuchs99:master 2015-03-17 11:12:49 +00:00
Vadim Pisarevsky
0f4d57d59e Merge pull request #3833 from sgrayb:houghline_bugfix 2015-03-17 10:37:10 +00:00
Vadim Pisarevsky
b708969326 Merge pull request #3832 from Dmitry-Me:fixNullPointerDereference3 2015-03-17 10:36:52 +00:00
Vadim Pisarevsky
8cf45ce0af Merge pull request #3828 from vpisarev:fix_win32_perf_calib3d_solvepnp_failure 2015-03-17 10:34:53 +00:00
Dmitry-Me
52a8d37f11 Fix potential null pointer dereference 2015-03-16 18:36:12 +03:00
Vadim Pisarevsky
ca19ae8b5a in solvePnPRansac call the solvePnP in the end with all the inliers to get more precise estimate 2015-03-16 16:56:26 +03:00
Vadim Pisarevsky
5c352c9146 temporarily use EPNP in SolvePnP instead of UPNP or DLS algorithms, since the latter two are not quite stable 2015-03-16 12:46:24 +03:00
Vadim Pisarevsky
3c54c50c5a Merge pull request #3802 from sanuj:master 2015-03-16 09:25:41 +00:00