Commit Graph

187 Commits

Author SHA1 Message Date
Ilya Lavrenov
fee3d6931b corners 2014-07-10 17:17:24 +04:00
Alexander Alekhin
e6c305cb75 Merge pull request #2897 from vbystricky:oclopt_sepFilter2D 2014-07-03 09:26:05 +00:00
Alexander Alekhin
acbabedac9 Merge pull request #2869 from ilya-lavrenov:tapi_warps 2014-06-30 11:59:13 +00:00
Alexander Alekhin
e0584bb818 Merge pull request #2876 from vbystricky:oclopt_integralsum 2014-06-30 10:06:54 +00:00
Alexander Alekhin
1493160f26 Merge pull request #2899 from ilya-lavrenov:tapi_ex 2014-06-30 10:05:24 +00:00
Alexander Alekhin
667a9328fb Merge pull request #2829 from ilya-lavrenov:tapi_cvtcolor 2014-06-30 10:02:41 +00:00
Ilya Lavrenov
36db85a94d optimized some operations 2014-06-27 12:52:29 +04:00
vbystricky
1a73aa1f6a Change local size 2014-06-26 15:43:40 +04:00
Ilya Lavrenov
16ab6ec534 ported changes from PR #2867 2014-06-26 13:52:44 +04:00
vbystricky
730ead44fe Optimize OpenCL version of sepFilter2D 2014-06-26 12:46:03 +04:00
Alexander Alekhin
ea2228774d Merge pull request #2881 from mlyashko:moments_opt 2014-06-24 16:19:52 +00:00
Alexander Alekhin
ea1b14ee95 Merge pull request #2786 from ElenaGvozdeva:ocl_matchTemplate 2014-06-24 16:12:42 +00:00
vbystricky
09bcc061dd Change kernel for optimization. Remove restriction to align data
Fix kernel compilation errors on AMD system

Fix licanse information in cl file

Support CV_64F destination type

Change build options of the kernel

Optimize sum of square

Remove separate kernel for integral square

Increase epsilon for perfomance tests

Increase epsilon for perfomance tests

Test double support on AMD devices

Fix some issues

Try to fix problems with AMD device

Try to solve problem with AMD device

Fix error of destination size in kernel

Fix warnings
2014-06-24 18:32:52 +04:00
Elena Gvozdeva
c23da52564 Fixed core for CCORR and SQDIFF. Used float instead of int for CV_8U. Fixed conditions for call dft. 2014-06-23 11:37:47 +04:00
mlyashko
44ffa42064 replaced factors computation by precomputed values, added kernel for
binary mode
2014-06-23 11:04:39 +04:00
Ilya Lavrenov
007593cab7 cvtColor - optimized index calculations; usage of build-in functions 2014-06-23 02:10:42 +04:00
Alexander Alekhin
6952b90ed0 Merge pull request #2790 from akarsakov:ocl_pyrUp_unroll 2014-06-20 13:49:14 +00:00
Alexander Alekhin
6e0d77469a Merge pull request #2864 from ilya-lavrenov:tapi_boxfilter 2014-06-20 08:47:51 +00:00
Alexander Alekhin
7688a18264 Merge pull request #2840 from ilya-lavrenov:tapi_calchist 2014-06-20 08:46:32 +00:00
Alexander Alekhin
ce5bae1c67 Merge pull request #2871 from vbystricky:oclopt_integral 2014-06-17 18:23:31 +04:00
Alexander Karsakov
06fb5da7c8 Removed storing of zeros in local memory 2014-06-17 17:24:11 +04:00
Elena Gvozdeva
dcaa8735ba used vector data types for CCORR cn==1 2014-06-17 17:09:39 +04:00
Elena Gvozdeva
13db948023 added dft for CCORR 2014-06-17 17:09:38 +04:00
Elena Gvozdeva
82da445a15 changed CCOEFF cn==1 2014-06-17 17:09:38 +04:00
Elena Gvozdeva
a7036d9668 changed support for 3-channels, changed CCOEFF 2014-06-17 17:06:51 +04:00
Alexander Alekhin
04628d770c Merge pull request #2849 from ElenaGvozdeva:ocl_matchTemplate_3cn 2014-06-17 14:37:52 +04:00
Ilya Lavrenov
93712e913a optimization of cv::warpAffine INTER_CUBIC 2014-06-17 00:47:19 +04:00
vbystricky
606df0469a Fix pointer conversion 2014-06-16 18:14:05 +04:00
vbystricky
9bf296eeb0 Small refactoring 2014-06-16 17:17:16 +04:00
Ilya Lavrenov
c424d36041 optimized cv::boxFilter 2014-06-16 17:00:10 +04:00
vbystricky
6550c4f682 Join kernel code for int and float destination types 2014-06-16 15:08:15 +04:00
vbystricky
504bc7634a Remove pre_invalid parameter 2014-06-16 13:07:39 +04:00
Alexander Alekhin
14671e0cd6 Merge pull request #2823 from ilya-lavrenov:tapi_warpaffine 2014-06-11 14:41:29 +04:00
Elena Gvozdeva
feeb386bf3 Added support for 3-channels 2014-06-11 11:49:22 +04:00
Ilya Lavrenov
33239fca70 cv::equalizeHist 2014-06-10 19:32:46 +04:00
Ilya Lavrenov
c9528b3952 optimized histogram merging 2014-06-10 19:32:01 +04:00
Ilya Lavrenov
eeaa4b3665 eliminated convertTo 2014-06-10 19:32:01 +04:00
Ilya Lavrenov
c072c28e28 optimized cv::calcHist 2014-06-10 19:32:00 +04:00
Alexander Alekhin
e430ab1a58 Merge pull request #2785 from akarsakov:ocl_pyrDown_borders 2014-06-06 18:53:14 +04:00
Alexander Alekhin
a42259421c Merge pull request #2794 from mletavin:pullreq/140516-median 2014-06-04 16:36:37 +04:00
Ilya Lavrenov
2cc4cf3644 optimized cv::warpAffine 2014-06-03 13:45:22 +04:00
Alexander Alekhin
a049c16ca7 Merge pull request #2751 from akarsakov:ocl_thresh_opt 2014-06-02 13:54:31 +04:00
mletavin
df9c75b270 Moved new kernels under conditional compilation to disable their build for 3-channel images; added condition to use new kernels only for images that are big enough 2014-05-27 16:16:00 +04:00
Alexander Karsakov
8e54845049 Removed useless multiplication by 4 2014-05-27 10:52:20 +04:00
mletavin
4a37ac303a Added new Intel-optimized 3x3 and 5x5 kernels to medianFilter.cl file and corresponding code to host in smooth.cpp
Only Intel platform and 1 channel images are supported.
2014-05-26 15:54:18 +04:00
Alexander Karsakov
5022a0fae3 Added new border types for pyrDown 2014-05-26 13:57:03 +04:00
Alexander Karsakov
4910242732 Unroll pyrUp kernel 2014-05-23 14:58:34 +04:00
Alexander Karsakov
8ecb8c6a3d Optimized memory access by using stride pattern 2014-05-16 10:22:03 +04:00
Aaron Kunze
033aebe668 A more efficient workaround for erode 2014-05-15 16:31:24 -07:00
Alexander Karsakov
4612b4b827 Added clamp() for THRESH_TRUNC mode 2014-05-13 15:21:47 +04:00