Alexander Alekhin [Fri, 27 Apr 2018 10:57:37 +0000 (10:57 +0000)]
Merge pull request #11410 from alalek:ipp_norm_L1_16UC3_mask_AVX512
Alexander Alekhin [Fri, 27 Apr 2018 09:56:42 +0000 (12:56 +0300)]
core: disabled IPP AVX512 normL1(a, b, mask)
for cv::Mat with type=16UC3 and width < 16
Alexander Alekhin [Thu, 26 Apr 2018 19:04:37 +0000 (19:04 +0000)]
Merge pull request #11402 from alalek:build_warnings
Alexander Alekhin [Thu, 26 Apr 2018 15:31:09 +0000 (15:31 +0000)]
Merge pull request #11403 from alalek:cmake_fix_cpu_baseline
Alexander Alekhin [Thu, 26 Apr 2018 15:29:35 +0000 (15:29 +0000)]
Merge pull request #11404 from alalek:icc_fix_std_exception_ptr
Alexander Alekhin [Thu, 26 Apr 2018 14:46:25 +0000 (17:46 +0300)]
core: fix icc std::exception_ptr detection
std::exception_ptr requires enabled C++11 mode
Alexander Alekhin [Thu, 26 Apr 2018 13:03:45 +0000 (16:03 +0300)]
cmake: fix CPU_BASELINE_FINAL filling
- remove duplicates
- restore "always on" missing entries
- fix FP16 detection on MSVC
Alexander Alekhin [Thu, 26 Apr 2018 12:13:19 +0000 (12:13 +0000)]
Merge pull request #11377 from Turim:master
Alexander Alekhin [Thu, 26 Apr 2018 12:06:59 +0000 (15:06 +0300)]
cuda: eliminate warnings
Alexander Alekhin [Thu, 26 Apr 2018 11:39:52 +0000 (11:39 +0000)]
Merge pull request #11400 from alalek:ipp_norm_L1_16UC3_mask
Alexander Alekhin [Thu, 26 Apr 2018 11:16:03 +0000 (14:16 +0300)]
videoio(openni2): fix CV_CAP_PROP_OPENNI2_SYNC property
Alexander Alekhin [Thu, 26 Apr 2018 11:13:01 +0000 (14:13 +0300)]
build: fix warnings
Alexander Enaldiev [Mon, 23 Apr 2018 12:15:34 +0000 (15:15 +0300)]
protobuf build documenting: move the WITH_PROTOBUF option to the top level
Alexander Alekhin [Thu, 26 Apr 2018 09:42:37 +0000 (12:42 +0300)]
core: disabled IPP AVX2 normL1(a, b, mask)
for cv::Mat with width < 16
Alexander Alekhin [Wed, 25 Apr 2018 16:32:53 +0000 (19:32 +0300)]
core: norm with mask 16UC3 regression test
Alexander Alekhin [Thu, 26 Apr 2018 10:31:31 +0000 (10:31 +0000)]
Merge pull request #11395 from alalek:fix_fixedpoint_build
Alexander Alekhin [Wed, 25 Apr 2018 19:42:13 +0000 (19:42 +0000)]
Merge pull request #11391 from alalek:android_pack_fix_contrib
Alexander Alekhin [Wed, 25 Apr 2018 14:18:15 +0000 (17:18 +0300)]
imgproc: fixed fixedpoint coding style
- fixed wrong condition (always true)
- replaced unnecessary shift operation: '>> 63' to '< 0'
- used CV_BIG_INT()/CV_BIG_UINT() macros
Alexander Alekhin [Wed, 25 Apr 2018 13:34:44 +0000 (16:34 +0300)]
imgproc: fix fixed point build
avoid using of templated 'operator T ()'
Alexander Alekhin [Wed, 25 Apr 2018 10:20:03 +0000 (13:20 +0300)]
android: don't use relative paths to contrib modules
Alexander Alekhin [Tue, 24 Apr 2018 15:11:14 +0000 (18:11 +0300)]
Merge pull request #11385 from shengyu7697:tab_to_space
Alexander Alekhin [Tue, 24 Apr 2018 15:06:59 +0000 (15:06 +0000)]
Merge pull request #11381 from alalek:replace_cv_errornoreturn
shengyu [Tue, 24 Apr 2018 12:34:47 +0000 (20:34 +0800)]
tab to space
Alexander Alekhin [Tue, 24 Apr 2018 11:29:01 +0000 (14:29 +0300)]
build: fix warnings
Alexander Alekhin [Tue, 24 Apr 2018 11:27:17 +0000 (14:27 +0300)]
build: -Wimplicit-fallthrough warning for GCC 7+ only
Alexander Alekhin [Tue, 24 Apr 2018 10:48:21 +0000 (13:48 +0300)]
build: unreachable code after CV_Error() (part 2)
Alexander Alekhin [Mon, 23 Apr 2018 16:02:39 +0000 (19:02 +0300)]
refactor: don't use CV_ErrorNoReturn() internally
Dmitry Kurtaev [Tue, 24 Apr 2018 11:59:59 +0000 (14:59 +0300)]
Custom layers for deep learning networks (#11129)
* Custom deep learning layers support
* Stack custom deep learning layers
Alexander Alekhin [Tue, 24 Apr 2018 11:44:33 +0000 (11:44 +0000)]
Merge pull request #11332 from alalek:v_select_x86
Vadim Pisarevsky [Tue, 24 Apr 2018 11:30:18 +0000 (11:30 +0000)]
Merge pull request #11380 from vpisarev:faster_interleave_deinterleave
Vadim Pisarevsky [Tue, 24 Apr 2018 11:29:50 +0000 (11:29 +0000)]
Merge pull request #11382 from alalek:cmake_macosx_cpu_baseline
Alexander Alekhin [Tue, 24 Apr 2018 09:40:25 +0000 (12:40 +0300)]
cmake: fix popcnt detection
via GCC/Clang __POPCNT__ define
Maksim Shabunin [Tue, 24 Apr 2018 08:23:56 +0000 (08:23 +0000)]
Merge pull request #11269 from allnes:gst_add_encoder_sample
Maksim Shabunin [Tue, 24 Apr 2018 08:23:28 +0000 (08:23 +0000)]
Merge pull request #11320 from mshabunin:gstreamer-cpp
Alexander Nesterov [Sun, 8 Apr 2018 22:27:55 +0000 (01:27 +0300)]
Added encode-pipeline to sample
Alexander Alekhin [Mon, 23 Apr 2018 16:42:49 +0000 (19:42 +0300)]
cmake: change CPU_BASELINE=DETECT for MacOSX
Vadim Pisarevsky [Mon, 23 Apr 2018 15:06:06 +0000 (18:06 +0300)]
improved performance of v_load_deinterleave(8uC3) & v_store_interleave(8uC3) intrinsics when using SSSE3 instructions.
Alexander Alekhin [Mon, 23 Apr 2018 14:51:55 +0000 (14:51 +0000)]
Merge pull request #11376 from alalek:cv_error_set_terminate
Alexander Alekhin [Mon, 23 Apr 2018 14:01:16 +0000 (14:01 +0000)]
Merge pull request #11375 from alalek:warnings_unreachable_code
Alexander Alekhin [Mon, 23 Apr 2018 11:32:40 +0000 (14:32 +0300)]
build: unreachable code after CV_Error()
Alexander Alekhin [Mon, 23 Apr 2018 12:06:43 +0000 (15:06 +0300)]
core: CV_Error with set_terminate() on Windows
To dump contents of the last OpenCV error
Maksim Shabunin [Mon, 16 Apr 2018 12:59:25 +0000 (15:59 +0300)]
VideoIO: prettier test console output, gstreamer capture backend changed to C++ interface
Alexander Alekhin [Tue, 17 Apr 2018 14:50:23 +0000 (17:50 +0300)]
core(hal): improve v_select() SSE4.1+
v_select 'mask' is restricted to these values only: 0 or ~0 (0xff/0xffff/etc)
mask in accuracy test is updated.
Vadim Pisarevsky [Mon, 23 Apr 2018 10:09:29 +0000 (10:09 +0000)]
Merge pull request #11358 from mshabunin:enable-xine
Vadim Pisarevsky [Mon, 23 Apr 2018 10:08:10 +0000 (10:08 +0000)]
Merge pull request #11350 from alalek:fix_11348
Vadim Pisarevsky [Mon, 23 Apr 2018 10:06:43 +0000 (10:06 +0000)]
Merge pull request #11357 from alalek:disable_errors_dump
Alexander Alekhin [Mon, 23 Apr 2018 09:24:39 +0000 (09:24 +0000)]
Merge pull request #11363 from mshabunin:compat-vc2017
Alexander Alekhin [Mon, 23 Apr 2018 09:16:39 +0000 (09:16 +0000)]
Merge pull request #11351 from dkurt:dnn_enable_inf_engine_tests
Alexander Alekhin [Mon, 23 Apr 2018 09:14:23 +0000 (09:14 +0000)]
Merge pull request #11374 from lupustr3:pvlasov/morph_big_fix
Pavel Vlasov [Mon, 23 Apr 2018 06:45:10 +0000 (09:45 +0300)]
IPP morphology with big images hotfix.
Maksim Shabunin [Fri, 20 Apr 2018 15:12:51 +0000 (18:12 +0300)]
cmake config: adding fallback to VS 2015 if distribution does not have VS 2017 binaries
Alexander Alekhin [Sat, 21 Apr 2018 19:11:21 +0000 (19:11 +0000)]
Merge pull request #11367 from berak:core_fix_inrange
berak [Sat, 21 Apr 2018 09:43:05 +0000 (11:43 +0200)]
core: add a check for empty input in inRange()
Alexander Alekhin [Fri, 20 Apr 2018 16:53:38 +0000 (16:53 +0000)]
Merge pull request #11362 from dkurt:dnn_keras_deconv
Ryan Wong [Fri, 20 Apr 2018 15:43:47 +0000 (08:43 -0700)]
Merge pull request #11304 from kinchungwong:issue_11242_intrin_cv34x_nocpp11
* Issue 11242 intrinsics v_extract, v_rotate improvement, branch 3.4, without C++11 (remove type restrictions for SSE2, use PALIGNR on SSSE3, compile to no-op when imm is 0 or nlanes).
* fix whitespace
* Fix #11242 (NEON intrinsics v_rotate...) branch 3.4
Separate macro expansion OPENCV_HAL_IMPL_NEON_SHIFT_OP for bitwise shifts for integers, from macro expansion OPENCV_HAL_IMPL_NEON_ROTATE for lane rotations. Bitwise shifts do not apply to floats, but lane-rotations can apply to both.
* fix whitespace
* Fix #11242 compile error (VSX intrinsics v_rotate(a)) branch 3.4 no-c++11
Alexander Alekhin [Fri, 20 Apr 2018 14:50:40 +0000 (14:50 +0000)]
Merge pull request #11315 from tomoaki0705:featureComparePixelCount
Alexander Alekhin [Fri, 20 Apr 2018 14:49:37 +0000 (14:49 +0000)]
Merge pull request #11361 from alalek:ocl_use_host_mem_ptr_workaround
Dmitry Kurtaev [Fri, 20 Apr 2018 13:44:05 +0000 (16:44 +0300)]
Fuse deconvolution layer subgraphs from Keras
Simon Que [Fri, 20 Apr 2018 12:31:47 +0000 (08:31 -0400)]
Merge pull request #11353 from eecsninja:3.4
* Fix CV_Asserts with negation of strings
{!"string"} causes some compilers to throw a warning.
The value of the string is not that important -- it's only for printing
the assertion message.
Replace these calls with:
CV_Error(Error::StsError, "string")
to suppress the warning.
* remove unnecessary 'break' after CV_Error()
Alexander Alekhin [Fri, 20 Apr 2018 12:30:50 +0000 (12:30 +0000)]
Merge pull request #11352 from alalek:build_implicit_fallthrough
Alexander Alekhin [Fri, 20 Apr 2018 12:29:13 +0000 (12:29 +0000)]
Merge pull request #11349 from alalek:cv_error_no_return
Maksim Shabunin [Fri, 20 Apr 2018 09:30:54 +0000 (12:30 +0300)]
videoio: refactored XINE backend
Alexander Alekhin [Fri, 20 Apr 2018 11:51:55 +0000 (14:51 +0300)]
ocl: CL_MEM_USE_HOST_PTR workaround test
Alexander Alekhin [Fri, 20 Apr 2018 09:29:12 +0000 (12:29 +0300)]
core: disable errors dump
Error messages are available via exceptions
Dmitry Kurtaev [Thu, 19 Apr 2018 12:04:57 +0000 (15:04 +0300)]
Enable some tests for clDNN plugin from Intel's Inference Engine
Alexander Alekhin [Thu, 19 Apr 2018 16:54:03 +0000 (19:54 +0300)]
build: enable -Wimplicit-fallthrough warning for OpenCV modules
Alexander Alekhin [Thu, 19 Apr 2018 15:42:01 +0000 (18:42 +0300)]
python: fix support of UI callbacks
Alexander Alekhin [Thu, 19 Apr 2018 13:14:27 +0000 (16:14 +0300)]
core: CV_Error is marked as 'noreturn'
Vadim Pisarevsky [Thu, 19 Apr 2018 12:23:50 +0000 (15:23 +0300)]
Grabcut with frozen models (#11339)
* model is not learned when grabcut is called with GC_EVAL
* fixed test, was writing to wrong file.
* modified patch by Iwan Paolucci; added GC_EVAL_FREEZE_MODEL in addition to GC_EVAL (which semantics is retained)
Vadim Pisarevsky [Thu, 19 Apr 2018 12:23:15 +0000 (12:23 +0000)]
Merge pull request #11340 from dkurt:dnn_inf_engine_switch_target
Vadim Pisarevsky [Thu, 19 Apr 2018 12:21:50 +0000 (12:21 +0000)]
Merge pull request #11341 from alalek:v_min_max_sse41
Alexander Alekhin [Thu, 19 Apr 2018 10:58:00 +0000 (10:58 +0000)]
Merge pull request #11309 from K-Shinotsuka:issue41
Tomoaki Teshima [Thu, 19 Apr 2018 08:31:42 +0000 (17:31 +0900)]
add new comparison to tolerate with rounding error
* restore resize of carotene
* clean up the source code
* remove unused member function (Near)
* add strict test on the border
Dmitry Kurtaev [Wed, 18 Apr 2018 14:26:54 +0000 (17:26 +0300)]
Let switch CPU/OpenCL targets for models from Intel's Model Optimizer
Alexander Alekhin [Wed, 18 Apr 2018 16:40:59 +0000 (19:40 +0300)]
core(hal): v_min()/v_max() use SSE4.1 intrinsics
Alexander Alekhin [Wed, 18 Apr 2018 13:34:34 +0000 (13:34 +0000)]
Merge pull request #11338 from alalek:fix_fp16_with_disabled_optimizations
Vadim Pisarevsky [Wed, 18 Apr 2018 12:11:12 +0000 (12:11 +0000)]
Merge pull request #11322 from dkurt:dnn_yolov3
Alexander Alekhin [Wed, 18 Apr 2018 10:37:22 +0000 (13:37 +0300)]
core: fix FP16 conversion with CV_DISABLE_OPTIMIZATION option
Reproducer:
cmake -DCPU_BASELINE=AVX2 -DCV_DISABLE_OPTIMIZATION=ON ...
Alexander Alekhin [Wed, 18 Apr 2018 09:09:08 +0000 (09:09 +0000)]
Merge pull request #11333 from alalek:cmake_fix_native
Alexander Alekhin [Tue, 17 Apr 2018 16:41:29 +0000 (16:41 +0000)]
Merge pull request #11331 from dkurt:fix_11318
Alexander Alekhin [Tue, 17 Apr 2018 15:58:07 +0000 (18:58 +0300)]
cmake: fix CPU_BASELINE=NATIVE on MSVS
Dmitry Kurtaev [Tue, 17 Apr 2018 14:53:51 +0000 (17:53 +0300)]
Fix bugs found by valgrind
Vadim Pisarevsky [Tue, 17 Apr 2018 12:50:52 +0000 (15:50 +0300)]
Rewite polar transforms (#11323)
* Rewrite polar transformations
- A new wrapPolar function encapsulate both linear and semi-log remap
- Destination size is a parameter or calculated automatically to keep objects size between remapping
- linearPolar and logPolar has been deprecated
* Fix build warning and error in accuracy test
* Fix function name to warpPolar
* Explicitly specify the mapping mode, so we retain all the parameters as non-optional.
Introduces WarpPolarMode enum to specify the mapping mode in flags
* resolves performance warning on windows build
* removed duplicated logPolar and linearPolar implementations
k-shinotsuka [Sat, 14 Apr 2018 14:11:01 +0000 (23:11 +0900)]
add universal intrinsics for RGB2HLS_f
Dmitry Kurtaev [Fri, 13 Apr 2018 15:53:12 +0000 (18:53 +0300)]
Support YOLOv3 model from Darknet
Alexander Alekhin [Mon, 16 Apr 2018 15:39:22 +0000 (15:39 +0000)]
Merge pull request #11319 from alalek:issue_11293
Alexander Alekhin [Mon, 16 Apr 2018 15:38:57 +0000 (15:38 +0000)]
Merge pull request #11317 from alalek:issue_11303
Alexander Alekhin [Mon, 16 Apr 2018 11:16:26 +0000 (14:16 +0300)]
dnn(test): update error tolerance for RCNN/RFCN tests
Alexander Alekhin [Mon, 16 Apr 2018 10:29:14 +0000 (13:29 +0300)]
imgproc(ipp): disable parallel GaussianBlur
Alexander Alekhin [Sat, 14 Apr 2018 11:36:19 +0000 (11:36 +0000)]
Merge pull request #11306 from alalek:opencl_trace_messages
Alexander Alekhin [Fri, 13 Apr 2018 13:24:06 +0000 (13:24 +0000)]
Merge pull request #11169 from tomoaki0705:universalRemap
Alexander Alekhin [Fri, 13 Apr 2018 12:56:41 +0000 (12:56 +0000)]
Merge pull request #11305 from tomoaki0705:typoNVIDIA
Tomoaki Teshima [Fri, 13 Apr 2018 12:19:16 +0000 (21:19 +0900)]
remove raw SSE2/NEON implementation from imgwarp.cpp
* use universal intrinsic instead of raw intrinsic
* add 2 channels de-interleave on x86 platform
* add v_int32x4 version of v_muladd
* add accumulate version of v_dotprod based on the commit from seiko2plus on bf1852d
* remove some verify check in performance test
* avoid the out of boundary access and keep the performance
Alexander Alekhin [Thu, 12 Apr 2018 16:28:46 +0000 (19:28 +0300)]
ocl: improve trace messages of OpenCL calls
Alexander Alekhin [Fri, 13 Apr 2018 11:49:38 +0000 (11:49 +0000)]
Merge pull request #11302 from dkurt:dnn_inf_engine_extra_layers
Tomoaki Teshima [Fri, 13 Apr 2018 11:33:19 +0000 (20:33 +0900)]
use correct name for NVIDIA
* remove NVidia and Nvidia
* replace Cuda with CUDA
* keep the letters for API
Dmitry Kurtaev [Thu, 12 Apr 2018 12:21:08 +0000 (15:21 +0300)]
Blank and L2-normalization layers from Intel's Inference Engine
Alexander Alekhin [Thu, 12 Apr 2018 10:57:39 +0000 (10:57 +0000)]
Merge pull request #11297 from seiko2plus:VSXImprovements_1
Sayed Adel [Wed, 11 Apr 2018 18:34:23 +0000 (18:34 +0000)]
core:ppc Several improvements on VSX(1)
* remove unnecessary defines from vsx_utils
* fix v_load_expand, load lower 64bit
* use vec_ld, vec_st with alignment load/store on all types except 64bit
* map v_extract to v_rotate_right
* update license header
* enable VSX by default on clang since #11167
Vadim Pisarevsky [Wed, 11 Apr 2018 15:25:24 +0000 (15:25 +0000)]
Merge pull request #11295 from dkurt:dnn_repeated_conv_params