Alexander Alekhin [Tue, 10 Sep 2019 09:57:38 +0000 (09:57 +0000)]
Merge pull request #15466 from alalek:ffmpeg_update_3.4
Alexander Alekhin [Fri, 6 Sep 2019 15:57:52 +0000 (15:57 +0000)]
Merge pull request #15475 from alalek:build_eliminate_warning_array_bounds
Alexander Alekhin [Fri, 6 Sep 2019 15:32:04 +0000 (15:32 +0000)]
Merge pull request #15461 from alalek:imgproc_connectedcomponents_allocation
Alexander Alekhin [Fri, 6 Sep 2019 14:48:24 +0000 (17:48 +0300)]
ts: eliminate -Warray-bounds warning
Alexander Alekhin [Thu, 5 Sep 2019 20:52:46 +0000 (20:52 +0000)]
ffmpeg/3.4: update FFmpeg wrapper
Alexander Alekhin [Thu, 5 Sep 2019 12:41:11 +0000 (15:41 +0300)]
imgproc: avoid manual memory allocation in connectedcomponents.cpp
Jaime [Thu, 5 Sep 2019 10:55:17 +0000 (06:55 -0400)]
Merge pull request #14872 from AhiyaHiya:feature/topmost_window
* Added mechanism to set/toggle NORMAL window to TOPMOST window.
* highgui: remove unnecessary/legacy code
Alexander Alekhin [Thu, 5 Sep 2019 10:44:06 +0000 (10:44 +0000)]
Merge pull request #15170 from sturkmen72:update_samples
Suleyman TURKMEN [Sun, 28 Jul 2019 09:09:17 +0000 (12:09 +0300)]
Update Samples
Alexander Alekhin [Wed, 4 Sep 2019 16:25:34 +0000 (16:25 +0000)]
Merge pull request #15444 from alalek:ocl_fix_fft_kernel
Everton Constantino [Wed, 4 Sep 2019 16:14:00 +0000 (13:14 -0300)]
Merge pull request #15440 from everton1984:new_integral_tests
* Adding all possible data type interactions to the perf tests since some
use SIMD acceleration and others do not.
* Disabling full tests by default.
* Giving proper names, removing magic numbers and sanity checks of new
performance tests for the integral function.
* Giving proper names, making array static.
huihut [Wed, 4 Sep 2019 15:36:56 +0000 (23:36 +0800)]
Merge pull request #15433 from huihut:master
* fix 'chmod' is not recognized as an internal or external command in Windows
https://github.com/opencv/opencv/issues/14939
* cmake: update PCH command handling
Dmitry Kurtaev [Tue, 3 Sep 2019 15:58:57 +0000 (16:58 +0100)]
Merge pull request #15063 from dkurt:dnn_ie_ocv_layers
* Wrap unsupported by IE layers as custom layers
* Replace pointers to layers blobs to their shapes
* Enable Faster R-CNN with IE backend on CPU
Alexander Alekhin [Tue, 3 Sep 2019 12:46:53 +0000 (15:46 +0300)]
core(ocl): fix fft kernel compilation
- error: variables in the local address space can only be declared in the outermost scope of a kernel function
Alexander Alekhin [Tue, 3 Sep 2019 12:21:02 +0000 (12:21 +0000)]
Merge pull request #15437 from devnexen:fbsd_opencl_build_fix
Alexander Alekhin [Tue, 3 Sep 2019 12:04:22 +0000 (12:04 +0000)]
Merge pull request #15435 from alalek:update_version_3.4.8-pre
Alexander Alekhin [Tue, 3 Sep 2019 10:50:45 +0000 (10:50 +0000)]
Merge pull request #15424 from mshabunin:add-cmake-docs
David Carlier [Mon, 2 Sep 2019 13:39:31 +0000 (13:39 +0000)]
OpenCL: FreeBSD build fix
Alexander Alekhin [Mon, 2 Sep 2019 11:20:49 +0000 (14:20 +0300)]
pre: OpenCV 3.4.8 (version++)
Alexander Alekhin [Mon, 2 Sep 2019 11:17:36 +0000 (14:17 +0300)]
experimental version++
Alexander Alekhin [Sat, 31 Aug 2019 11:16:04 +0000 (11:16 +0000)]
Merge pull request #15339 from pmur:dotprod-32s-vsx
Alexander Alekhin [Sat, 31 Aug 2019 11:10:05 +0000 (11:10 +0000)]
Merge pull request #15402 from ChipKerchner:normUnroll
Chip Kerchner [Sat, 31 Aug 2019 10:47:58 +0000 (06:47 -0400)]
Merge pull request #15358 from ChipKerchner:imgwarpToHal
* Convert ImgWarp from SSE SIMD to HAL - 2.8x faster on Power (VSX) and 15% speedup on x86
* Change compile flag from CV_SIMD128 to CV_SIMD128_64F for use of v_float64x2 type
* Changing WarpPerspectiveLine from class functions and dispatching to static functions.
* Re-add dynamic runtime and dispatch execution.
* RRestore SSE4_1 optimizations inside opt_SSE4_1 namespace
Maksim Shabunin [Fri, 30 Aug 2019 22:47:04 +0000 (01:47 +0300)]
Assorted documentation fixes
* removed private flann documentation
* common tutorial images moved to doc/images
* grouping issues
Alexander Alekhin [Fri, 30 Aug 2019 12:03:19 +0000 (12:03 +0000)]
Merge pull request #15414 from kuzi117:instr
Alexander Alekhin [Thu, 29 Aug 2019 18:42:58 +0000 (18:42 +0000)]
Merge pull request #15418 from mshabunin:try-install-layout-34
Alexander Alekhin [Thu, 29 Aug 2019 18:40:19 +0000 (18:40 +0000)]
Merge pull request #15417 from alalek:ts_fix_callback_signature
Alexander Alekhin [Thu, 29 Aug 2019 18:39:00 +0000 (18:39 +0000)]
Merge pull request #15412 from kuzi117:ambig
Braedy Kuzma [Thu, 29 Aug 2019 00:20:09 +0000 (18:20 -0600)]
Use commonly supported instruction mnemonic.
Braedy Kuzma [Wed, 28 Aug 2019 23:39:08 +0000 (17:39 -0600)]
Disambiguate vecpopcnt for (u)dword2.
Maksim Shabunin [Thu, 29 Aug 2019 13:57:42 +0000 (16:57 +0300)]
Ported install layout refactoring from master branch
Alexander Alekhin [Thu, 29 Aug 2019 11:37:35 +0000 (14:37 +0300)]
ts: fix callback function signature
detected by UBSAN
Alexander Alekhin [Thu, 29 Aug 2019 08:26:31 +0000 (08:26 +0000)]
Merge pull request #15357 from ChipKerchner:fastCorner
Lubov Batanina [Thu, 29 Aug 2019 07:52:02 +0000 (10:52 +0300)]
Merge pull request #15395 from l-bat:fully_connected
* Fix IE FullyConnected layer
* Fix MyriadX
Alexander Alekhin [Wed, 28 Aug 2019 16:39:01 +0000 (16:39 +0000)]
Merge pull request #15407 from terfendail:test_check_any
Chip Kerchner [Wed, 28 Aug 2019 15:56:48 +0000 (11:56 -0400)]
Merge pull request #15274 from ChipKerchner:lkpyramidToHal
* Convert lkpyramid from SSE SIMD to HAL - 90% faster on Power (VSX).
* Replace stores with reduce_sum. Rework to handle endianess correctly.
* Fix compiler warnings by casting values explicitly to shorts
* Switch to CV_SIMD128 compiler definition. Unroll loop to 8 elements since we've already loaded the data.
Vitaly Tuzov [Wed, 28 Aug 2019 11:53:31 +0000 (14:53 +0300)]
Extend tests for v_check_any and v_check_all intrinsics
Alexander Alekhin [Tue, 27 Aug 2019 19:59:39 +0000 (19:59 +0000)]
Merge pull request #15401 from ChipKerchner:vectorReduceInt8Bug
ChipKerchner [Tue, 27 Aug 2019 17:15:19 +0000 (12:15 -0500)]
\96Improve vectorization in the 'norm' functions
ChipKerchner [Tue, 27 Aug 2019 16:38:53 +0000 (11:38 -0500)]
Fix macro bug with v_reduce_min and v_reduce_max for chars in VSX
Alexander Alekhin [Tue, 27 Aug 2019 14:07:23 +0000 (14:07 +0000)]
Merge pull request #15400 from terfendail:fix_check_any
Vitaly Tuzov [Tue, 27 Aug 2019 11:31:23 +0000 (14:31 +0300)]
Fix for AVX2 implementation of v_check_any(), v_check_all() intrinsics
atinfinity [Mon, 26 Aug 2019 14:55:10 +0000 (23:55 +0900)]
Merge pull request #15388 from atinfinity:impl-turbo-colormap
Implementation of colormap "Turbo" (#15388)
* implemented turbo colormap
* add colormap image
* changed float value to avoid cast
* sorted flag check alphabetically
Alexander Alekhin [Mon, 26 Aug 2019 12:17:21 +0000 (12:17 +0000)]
Merge pull request #15391 from dkurt:fix_15141
Dmitry Kurtaev [Sat, 24 Aug 2019 20:14:26 +0000 (23:14 +0300)]
Fix https://github.com/opencv/opencv/issues/15141
Alexander Alekhin [Fri, 23 Aug 2019 16:19:36 +0000 (16:19 +0000)]
Merge pull request #15383 from alalek:fix_15379
Alexander Alekhin [Fri, 23 Aug 2019 16:09:13 +0000 (16:09 +0000)]
Merge pull request #15382 from alalek:fix_15287
Alexander Alekhin [Fri, 23 Aug 2019 16:07:21 +0000 (16:07 +0000)]
Merge pull request #15380 from alalek:fix_build_icc
Alexander Alekhin [Fri, 23 Aug 2019 16:05:15 +0000 (16:05 +0000)]
Merge pull request #15376 from ilya-lavrenov:detect_openvino
Giles Payne [Fri, 23 Aug 2019 15:22:29 +0000 (00:22 +0900)]
Merge pull request #15350 from komakai:apple-debug-build
* Add debug build option to MacOs and iOS build scripts
* osx: allow selection of MACOSX_DEPLOYMENT_TARGET
* Add --debug_info flag to iOS and macOS builds for building with debug info
Ilya Lavrenov [Fri, 23 Aug 2019 09:54:07 +0000 (12:54 +0300)]
Adopted detection of Inference Engine
Alexander Alekhin [Fri, 23 Aug 2019 13:54:24 +0000 (16:54 +0300)]
core: handle empty Mat in Mat_ assignment operators
Alexander Alekhin [Thu, 22 Aug 2019 13:26:58 +0000 (16:26 +0300)]
build: fix build with ICC
Alexander Alekhin [Fri, 23 Aug 2019 13:14:53 +0000 (16:14 +0300)]
objdetect: add input check in HOG detector
Alexander Alekhin [Thu, 22 Aug 2019 20:54:34 +0000 (20:54 +0000)]
Merge pull request #15373 from pmur:libpng-vsx-fix
Alexander Alekhin [Thu, 22 Aug 2019 20:52:54 +0000 (20:52 +0000)]
Merge pull request #15372 from alalek:core_stat_fix_intrin
Alexander Alekhin [Thu, 22 Aug 2019 17:05:38 +0000 (17:05 +0000)]
Merge pull request #15370 from alalek:core_fastmath_hpp_update
Alexander Alekhin [Thu, 22 Aug 2019 15:43:34 +0000 (15:43 +0000)]
Merge pull request #15353 from alalek:cmake_check_atomic
Paul E. Murphy [Thu, 22 Aug 2019 15:08:01 +0000 (10:08 -0500)]
libpng: update check for VSX enablement
The deprecated flag ENABLE_VSX should not be used to enable this
feature. Instead, use the baseline cpu feature set to determine.
Alexander Alekhin [Wed, 21 Aug 2019 22:21:53 +0000 (22:21 +0000)]
core: update fastmath.hpp
Alexander Alekhin [Thu, 22 Aug 2019 12:39:51 +0000 (15:39 +0300)]
core: fix stat SIMD code
Alexander Alekhin [Thu, 22 Aug 2019 11:37:38 +0000 (11:37 +0000)]
Merge pull request #15316 from sturkmen72:update_seamless_cloning_cpp
Prakhar Varshney [Thu, 22 Aug 2019 09:25:27 +0000 (14:55 +0530)]
Merge pull request #15345 from prakhar9998:patch-1
* fix broken code link
resolves #15344
* Update stitcher.markdown
* update 3.4 link
Alexander Alekhin [Wed, 21 Aug 2019 16:59:20 +0000 (16:59 +0000)]
Merge pull request #15303 from dkurt:fix_15296
Alexander Alekhin [Wed, 21 Aug 2019 15:55:25 +0000 (15:55 +0000)]
Merge pull request #15365 from Zyrin:3.4
Alexander Alekhin [Wed, 21 Aug 2019 15:10:25 +0000 (15:10 +0000)]
Merge pull request #15368 from dab0bby:patch-2
dab0bby [Wed, 21 Aug 2019 12:52:31 +0000 (14:52 +0200)]
fix typo and reference
Alexander Alekhin [Tue, 20 Aug 2019 13:19:37 +0000 (16:19 +0300)]
cmake: add libatomic check
Rosen Penev [Wed, 21 Aug 2019 09:59:09 +0000 (02:59 -0700)]
Merge pull request #15279 from neheb:patch-1
* jas_stream: Add definition for L_tmpnam if missing
uClibc-ng for some reason does not provide a definition when some
deprecated APIs are disabled. Value taken from the musl libc.
* 3rdparty: move uClibc-ng workaround into config file
Zyrin [Wed, 21 Aug 2019 09:12:00 +0000 (11:12 +0200)]
Use std::move in Mat_<T> move constructors
Zyrin [Wed, 21 Aug 2019 08:57:03 +0000 (10:57 +0200)]
Fix stack overflow on gcc with c++17 (#15343)
Alexander Alekhin [Tue, 20 Aug 2019 22:10:10 +0000 (22:10 +0000)]
Merge pull request #15335 from lixit:master
Paul E. Murphy [Tue, 20 Aug 2019 16:26:38 +0000 (11:26 -0500)]
core: vectorize dotProd_32s
Use 4x FMA chains to sum on SIMD 128 FP64 targets. On
x86 this showed about 1.4x improvement.
For PPC, do a full multiply (32x32->64b), convert to DP
then accumulate. This may be slightly less precise for
some inputs. But is 1.5x faster than the above which
is about 1.5x than the FMA above for ~2.5x speedup.
ChipKerchner [Tue, 20 Aug 2019 18:25:35 +0000 (13:25 -0500)]
Change fast corner flags in HAL version from char array to single int
xitong [Mon, 19 Aug 2019 08:08:53 +0000 (16:08 +0800)]
Fix a typo
Alexander Alekhin [Mon, 19 Aug 2019 19:01:43 +0000 (19:01 +0000)]
Merge pull request #15329 from craterkamath:fixTypoCamshift
craterkamath [Sun, 18 Aug 2019 20:36:31 +0000 (02:06 +0530)]
Fixed Typo Closes #15148
Alexander Alekhin [Fri, 16 Aug 2019 15:40:21 +0000 (15:40 +0000)]
Merge pull request #15322 from alalek:backport_15318
Suleyman TURKMEN [Thu, 15 Aug 2019 18:49:02 +0000 (21:49 +0300)]
Update seamless_cloning.cpp
luz.paz [Thu, 15 Aug 2019 22:02:09 +0000 (18:02 -0400)]
Fix modules/ typos
Found using `codespell -q 3 -S ./3rdparty -L activ,amin,ang,atleast,childs,dof,endwhile,halfs,hist,iff,nd,od,uint`
backporting of commit:
ec43292e1ea9da963e67427505b4113750829c3e
Alexander Alekhin [Thu, 15 Aug 2019 12:02:24 +0000 (12:02 +0000)]
Merge pull request #15309 from alalek:backport_15305
luz.paz [Wed, 14 Aug 2019 17:33:49 +0000 (13:33 -0400)]
FIx misc. source and comment typos
Found via `codespell -q 3 -S ./3rdparty,./modules -L amin,ang,atleast,dof,endwhile,hist,uint`
backporting of commit:
32aba5e64ba40b372f02b5fab18c9632f763bb75
Alexander Alekhin [Wed, 14 Aug 2019 19:28:05 +0000 (19:28 +0000)]
Merge pull request #15122 from pmur:fast-math-improvements
Alexander Alekhin [Wed, 14 Aug 2019 19:17:07 +0000 (19:17 +0000)]
Merge pull request #15101 from alalek:cmake_initialization
Dmitry Kurtaev [Wed, 14 Aug 2019 16:44:05 +0000 (19:44 +0300)]
Fix #15296
Alexander Alekhin [Wed, 14 Aug 2019 16:40:24 +0000 (16:40 +0000)]
Merge pull request #15301 from alalek:backport
Alexander Alekhin [Thu, 18 Jul 2019 19:40:23 +0000 (19:40 +0000)]
3.4: backported changes from 'master' branch
Alexander Alekhin [Tue, 13 Aug 2019 14:18:47 +0000 (14:18 +0000)]
Merge pull request #15292 from alalek:build_warnings_xcode_10_3
Alexander Alekhin [Tue, 13 Aug 2019 12:02:05 +0000 (15:02 +0300)]
build: eliminate warnings with Xcode 10.3
Alexander Alekhin [Mon, 12 Aug 2019 16:32:38 +0000 (16:32 +0000)]
Merge pull request #15249 from dkurt:dnn_ie_tests_2019r2
Hugo Lindström [Fri, 9 Aug 2019 15:01:37 +0000 (17:01 +0200)]
Merge pull request #15265 from hugolm84:wince-armv7-supports-neon
* WINCE 8.0 requires ARMv7 Thumb2 and thus have NEON instructions
* Only add NEON if on _ARM_
Chip Kerchner [Thu, 8 Aug 2019 15:57:03 +0000 (11:57 -0400)]
Merge pull request #15199 from ChipKerchner:hogToHal
Convert HOG from SSE SIMD to HAL - 35-45% faster on Power (VSX) (#15199)
* Convert SSE SIMD to HAL. 35-45% improvement for Power (VSX)
* Remove CV_NEON code. Use v_floor instead of 3 lines of code.
* Invert comparison logic to simplify code.
* Change initialization from v_load to constructor type.
Alexander Alekhin [Thu, 8 Aug 2019 15:47:08 +0000 (15:47 +0000)]
Merge pull request #15263 from alalek:videoio_skip_msmf_tests3
Alexander Alekhin [Thu, 8 Aug 2019 15:45:29 +0000 (15:45 +0000)]
Merge pull request #15260 from tomoaki0705:suppressNoisyWarning
Jake Howard [Thu, 8 Aug 2019 13:56:49 +0000 (14:56 +0100)]
Merge pull request #15111 from RealOrangeOne:patch-1
* Remove unavoidable print of CV error
The return value covers whether the device exists.
This might be better hidden behind a debug flag, but I couldn't work out how to do that nicely.
* Use `CV_LOG_WARNING` macro to log rather than removing it entirely
Alexander Alekhin [Thu, 8 Aug 2019 13:25:19 +0000 (16:25 +0300)]
videoio(test): skip unstable MSMF tests (3)
Tomoaki Teshima [Thu, 8 Aug 2019 12:49:32 +0000 (21:49 +0900)]
suppress noisy warning
* add -Wno-psabi when using GCC 6
* add -Wundef for CUDA 10
* add -Wdeprecated-declarations when using GCC 7
* add -Wstrict-aliasing and -Wtautological-compare for GCC 7
* replace cudaThreadSynchronize with cudaDeviceSynchronize
Alexander Alekhin [Thu, 18 Jul 2019 10:22:02 +0000 (13:22 +0300)]
cmake: update initialization
Paul E. Murphy [Mon, 22 Jul 2019 19:23:56 +0000 (14:23 -0500)]
fast_math: implement optimized PPC routines
Implement cvRound using inline asm. No compiler support
exists today to properly optimize this. This results in
about a 4x speedup over the default rounding. Likewise,
simplify the growing number of rounding function overloads.
For P9 enabled targets, utilize the classification
testing instruction to test for Inf/Nan values. Operation
speedup is about 1.2x for FP32, and 1.5x for FP64 operands.
For P8 targets, fallback to the GCC nan inline. It provides
a 1.1/1.4x improvement for FP32/FP64 arguments.
Paul E. Murphy [Mon, 22 Jul 2019 19:23:49 +0000 (14:23 -0500)]
fast_math: selectively use GCC rounding builtins when available
Add a new macro definition OPENCV_USE_FASTMATH_GCC_BUILTINS to enable
usage of GCC inline math functions, if available and requested by the
user.
Likewise, enable it for POWER. This is nearly always a substantial
improvement over using integer manipulation as most operations can
be done in several instructions with no branching. The result is a
1.5-1.8x speedup in the ceil/floor operations.
1. As tested with AT 12.0-1 (GCC 8.3.1) compiler on P9 LE.