Vadim Pisarevsky
|
0e46b4cd03
|
Merge pull request #3102 from akarsakov:ocl_denoising_fix
|
10 years ago |
Vadim Pisarevsky
|
75487d43f5
|
fixed trailing whitespaces
|
10 years ago |
Vadim Pisarevsky
|
9a5aa4b23c
|
trying to solve compile problems; temporarily disabled some tests for just added optimization algo's
|
10 years ago |
Alexander Alekhin
|
cf4a536440
|
Merge pull request #3098 from jet47:fix-cuda-compilation
|
10 years ago |
Vadim Pisarevsky
|
00b2124876
|
fixing more compile problems with Java bindings etc
|
10 years ago |
Vadim Pisarevsky
|
4530c7ad08
|
trying to fix builds
|
10 years ago |
Vadim Pisarevsky
|
f7243be79c
|
Merge pull request #3090 from akarsakov:ocl_reduce_fix
|
10 years ago |
Vadim Pisarevsky
|
5240a03217
|
Merge pull request #3100 from vpisarev:disabled_neon_split_merge_on_ios
|
10 years ago |
Vadim Pisarevsky
|
257463719b
|
removed optim module; moved its functionality to core and photo modules; moved drawing functions from core to imgproc. Removed FilterEngine etc. from public API
|
10 years ago |
vbystricky
|
aab6f6c56f
|
Set sync=true in ocl_minMaxIdx kernel call. For tests
|
10 years ago |
Alexander Karsakov
|
75ede6e17c
|
Minimal image size is set to 5*5 for OCL path of fastNlMeansDenoising and fastNlMeansDenoisingColored since it breaks tests with test_loop_times=30 and in general doesn't make sense for computation on GPU.
|
10 years ago |
Alexander Karsakov
|
8760a344ec
|
Fixed incorrect pointer conversion for in-place threshold
|
10 years ago |
Vadim Pisarevsky
|
6f86c9f118
|
another mac-related fix, let's squeeze it in this PR too
|
10 years ago |
Vadim Pisarevsky
|
5d8098a2c3
|
avoid compiler warnings
|
10 years ago |
VBystricky
|
c5632d412e
|
Change epsilon for NORM_INF_2args accuracy tests
|
10 years ago |
Vadim Pisarevsky
|
a602185fb6
|
Merge pull request #3097 from mshabunin:gdal-support
|
10 years ago |
Vadim Pisarevsky
|
5f590ebed0
|
Merge pull request #3094 from Adil-Ibragimov:minor_fixes_in_Mat-data
|
10 years ago |
Vadim Pisarevsky
|
71ff228ad3
|
Merge pull request #3089 from Adil-Ibragimov:fix_cuda_constness
|
10 years ago |
Vadim Pisarevsky
|
a37876ab95
|
fixed build of iOS framework another time; clang does not like the new NEON-optimized split/merge
|
10 years ago |
jaco
|
b81dcbdf2a
|
#pragma warning4127 fixation modified
Now use cleaner approach:
#pragma warning( push )
#pragma warning( disable: ThatWarning )
//code with ThatWarning here
#pragma warning( pop )
|
10 years ago |
Vladislav Vinogradov
|
9786e0a760
|
fix cudabgsegm module compilation
|
10 years ago |
vbystricky
|
1fe403f461
|
Enable OpenCL version of norm and convertScaleAbs or 32F data
Fix error in minmaxloc.cl
Change test for convertScaleAbs
Fix minMaxIdx for _src2 align
Change epsilon on the tests
|
10 years ago |
mshabunin
|
3ed89a8879
|
Merge with master and fix conflicts
|
10 years ago |
Vladislav Vinogradov
|
8392296eda
|
remove -fvisibility-inlines-hidden from CUDA compiler flags
it generates warning
|
10 years ago |
Vladislav Vinogradov
|
f5d9359da8
|
use directory-based include_directories command if we build with CUDA
|
10 years ago |
Vladislav Vinogradov
|
0825120f6a
|
move CUDA compilation step from ocv_glob_module_sources to ocv_add_library
|
10 years ago |
Vadim Pisarevsky
|
12c69ad490
|
Merge pull request #3095 from vpisarev:restore_ocl_div_mul_and_repeat
|
10 years ago |
Adil Ibragimov
|
9ac06a9333
|
fix for gcc
|
10 years ago |
Vadim Pisarevsky
|
d97f9fc477
|
now the tests pass with loop_times==30
|
10 years ago |
Adil Ibragimov
|
8a4a1bb018
|
Several type of formal refactoring:
1. someMatrix.data -> someMatrix.prt()
2. someMatrix.data + someMatrix.step * lineIndex -> someMatrix.ptr( lineIndex )
3. (SomeType*) someMatrix.data -> someMatrix.ptr<SomeType>()
4. someMatrix.data -> !someMatrix.empty() ( or !someMatrix.data -> someMatrix.empty() ) in logical expressions
|
10 years ago |
Vadim Pisarevsky
|
30111a786a
|
Merge pull request #3085 from nickdademo:pvapi_fix_decimation_enum
|
10 years ago |
Vadim Pisarevsky
|
3cd7941098
|
Merge pull request #3083 from akarsakov:ocl_transpose_check_locmem
|
10 years ago |
Vadim Pisarevsky
|
01bd4961e3
|
Merge pull request #3081 from akarsakov:small_fixes
|
10 years ago |
Alexander Karsakov
|
c3100eeb19
|
Fixed buffer initialization in reduce kernel. Enabled OCL version of reduce for SUM, MAX, MIN modes.
|
10 years ago |
Adil Ibragimov
|
dfd0b297ce
|
datastart must be non-const in CUDA.
|
10 years ago |
Alexander Alekhin
|
28f3a44e33
|
Merge pull request #3080 from ElenaGvozdeva:ocl_stich
|
10 years ago |
Alexander Karsakov
|
713ddb89bd
|
Set minimum matrix size for AmdBlas::gemm to 20 since it works incorrect for small sizes
|
10 years ago |
Vadim Pisarevsky
|
c97c0be4f2
|
Merge pull request #3082 from ElenaGvozdeva:ocl_morphSmall
|
10 years ago |
Vadim Pisarevsky
|
5505aa28ce
|
Merge pull request #3078 from vpisarev:refactor_features2d_take3
|
10 years ago |
jaco
|
58b3bc2d7e
|
#pragma warning4127 fixed for cvstd.inl.hpp
|
10 years ago |
Nick D'Ademo
|
91676de638
|
Change decimation enum values so they can be used in resolution calculations.
BUGFIX: Fixed bugged when setting PVAPI property CV_CAP_PROP_FRAME_HEIGHT.
|
10 years ago |
Alexander Karsakov
|
b4d3b34acf
|
Disabled ippiFilterBoxBorder_* in case maskSize equal roiSize for any dimension
|
10 years ago |
Elena Gvozdeva
|
1c7ee88b84
|
fixed test for ocl PlaneWarperTest
|
10 years ago |
Alexander Karsakov
|
3deea1c3ce
|
Added check required local memory size for non-inplace transpose kernel
|
10 years ago |
Vadim Pisarevsky
|
5119564579
|
Merge pull request #3033 from nathanjackson:clahe_datatypes
|
10 years ago |
Vadim Pisarevsky
|
1629d50999
|
removed some more trailing whitespaces
|
10 years ago |
Alexander Karsakov
|
5898dcae4a
|
Added ROUNDING_EPS for identical rounding after dividing on different platforms
|
10 years ago |
Alexander Alekhin
|
da806cc3e7
|
Merge pull request #3073 from vbystricky:ocl_fixMinMaxLoc
|
10 years ago |
Elena Gvozdeva
|
8124d10526
|
fixed Erode and Dilate in case of kernel = Mat()
|
10 years ago |
vbystricky
|
6fb282aa39
|
Remove mul24, for CV8UC3 3840x2160 it generates implementation specific result
|
10 years ago |