platform/upstream/opencv.git
7 years agofix build error on Visual Studio 2012
Tomoaki Teshima [Thu, 20 Jul 2017 13:55:10 +0000 (22:55 +0900)]
fix build error on Visual Studio 2012

7 years agoMerge pull request #9189 from tomoaki0705:fixCalib3dRandom
Alexander Alekhin [Thu, 20 Jul 2017 12:24:34 +0000 (12:24 +0000)]
Merge pull request #9189 from tomoaki0705:fixCalib3dRandom

7 years agoMerge pull request #9200 from alalek:perf_stitching_win32
Alexander Alekhin [Thu, 20 Jul 2017 12:08:35 +0000 (12:08 +0000)]
Merge pull request #9200 from alalek:perf_stitching_win32

7 years agoperf: skip stitching OpenCL test on Win32 platform
Alexander Alekhin [Thu, 20 Jul 2017 11:14:22 +0000 (14:14 +0300)]
perf: skip stitching OpenCL test on Win32 platform

7 years agofix the test fail on Calib3d_SolvePnP.accuracy
Tomoaki Teshima [Wed, 19 Jul 2017 21:32:32 +0000 (06:32 +0900)]
fix the test fail on Calib3d_SolvePnP.accuracy
  * move array size to enum
  * move array size to member variable
  * loosen the eps of SOLVEPNP_P3P
  * loosen the eps in Calib3d_SolveP3P.accuracy

7 years agoMerge pull request #9143 from sovrasov:gen_pattern_fix
Alexander Alekhin [Wed, 19 Jul 2017 15:46:42 +0000 (15:46 +0000)]
Merge pull request #9143 from sovrasov:gen_pattern_fix

7 years agoMerge pull request #9161 from alalek:separate_debug_symbols
Alexander Alekhin [Wed, 19 Jul 2017 15:34:43 +0000 (15:34 +0000)]
Merge pull request #9161 from alalek:separate_debug_symbols

7 years agoMerge pull request #9186 from alalek:cmake_fix_dump_duplication
Alexander Alekhin [Wed, 19 Jul 2017 11:21:47 +0000 (11:21 +0000)]
Merge pull request #9186 from alalek:cmake_fix_dump_duplication

7 years agocmake: ocv_cmake_dump_vars() remove duplicated information
Alexander Alekhin [Wed, 19 Jul 2017 09:37:36 +0000 (12:37 +0300)]
cmake: ocv_cmake_dump_vars() remove duplicated information

7 years agoMerge pull request #9037 from arrybn:googlenet_test
Alexander Alekhin [Tue, 18 Jul 2017 16:35:00 +0000 (16:35 +0000)]
Merge pull request #9037 from arrybn:googlenet_test

7 years agoMerge pull request #9181 from alalek:stitching_perf_test
Alexander Alekhin [Tue, 18 Jul 2017 15:53:32 +0000 (15:53 +0000)]
Merge pull request #9181 from alalek:stitching_perf_test

7 years agoMerge pull request #9180 from alalek:fix_photo_crash_win32
Alexander Alekhin [Tue, 18 Jul 2017 15:52:25 +0000 (15:52 +0000)]
Merge pull request #9180 from alalek:fix_photo_crash_win32

7 years agoRewrote googlenet tests
Aleksandr Rybnikov [Thu, 29 Jun 2017 13:45:17 +0000 (16:45 +0300)]
Rewrote googlenet tests

7 years agostitching(perf): check for available OpenCL memory
Alexander Alekhin [Tue, 18 Jul 2017 14:25:35 +0000 (17:25 +0300)]
stitching(perf): check for available OpenCL memory

7 years agophoto: crash workaround for MSVC 2015 32-bit
Alexander Alekhin [Tue, 18 Jul 2017 13:04:35 +0000 (16:04 +0300)]
photo: crash workaround for MSVC 2015 32-bit

7 years agoMerge pull request #8967 from savuor:rgb2lab_faster
Alexander Alekhin [Mon, 17 Jul 2017 19:55:19 +0000 (19:55 +0000)]
Merge pull request #8967 from savuor:rgb2lab_faster

7 years agoRGB2Lab_f and trilinear interpolation code are in separate branch; cubeRoot(x) instea...
Rostislav Vasilikhin [Fri, 20 Jan 2017 18:56:44 +0000 (21:56 +0300)]
RGB2Lab_f and trilinear interpolation code are in separate branch; cubeRoot(x) instead of std::pow(x, 1.f/3.f)

file with internal accuracy&speed tests moved to lab_tetra branch

7 years agoMerge pull request #8498 from savuor:bit_exact_lab
Alexander Alekhin [Mon, 17 Jul 2017 14:01:05 +0000 (14:01 +0000)]
Merge pull request #8498 from savuor:bit_exact_lab

7 years agoMerge pull request #9176 from alalek:fix_nightly_builds
Alexander Alekhin [Mon, 17 Jul 2017 13:37:46 +0000 (13:37 +0000)]
Merge pull request #9176 from alalek:fix_nightly_builds

7 years agoMerge pull request #9175 from alalek:issue_9169
Alexander Alekhin [Mon, 17 Jul 2017 13:05:11 +0000 (13:05 +0000)]
Merge pull request #9175 from alalek:issue_9169

7 years agoimgproc: fix warp optimizations
Alexander Alekhin [Mon, 17 Jul 2017 12:12:41 +0000 (15:12 +0300)]
imgproc: fix warp optimizations

7 years agocore: fix convertTo() AVX2 optimization
Alexander Alekhin [Mon, 17 Jul 2017 12:02:14 +0000 (15:02 +0300)]
core: fix convertTo() AVX2 optimization

7 years agocore: clarify documentation of cv::Mat::deallocate() method
Alexander Alekhin [Mon, 17 Jul 2017 10:31:47 +0000 (13:31 +0300)]
core: clarify documentation of cv::Mat::deallocate() method

7 years agoMerge pull request #9173 from thcd:patch-1
Alexander Alekhin [Mon, 17 Jul 2017 10:09:33 +0000 (10:09 +0000)]
Merge pull request #9173 from thcd:patch-1

7 years agoMerge pull request #9172 from atinfinity:fixed_cxx11_flag
Alexander Alekhin [Mon, 17 Jul 2017 07:56:30 +0000 (07:56 +0000)]
Merge pull request #9172 from atinfinity:fixed_cxx11_flag

7 years agoFix typo in imgproc.hpp
Shuyu Liang [Mon, 17 Jul 2017 07:51:10 +0000 (15:51 +0800)]
Fix typo in imgproc.hpp

7 years agoinitial version of Lab2RGB_f tetrahedral interpolation written
Rostislav Vasilikhin [Fri, 20 Jan 2017 18:56:44 +0000 (21:56 +0300)]
initial version of Lab2RGB_f tetrahedral interpolation written

RGB2Lab_f added, bugs fixed, moved to float

several bugs fixed

LUT fixed, no switch in tetraInterpolate()

temporary code; to be removed and rewritten

before refactoring

extra interpolations removed, some things to do left

added Lab2RGB_b +XYZ version, etc.

basic version is done, to be sped up

tetra refactored

interpolations: LUT for weights, refactor., etc.

address arithm optimized

initial version of vectorized code added (not compiling now)

compilation fixed, now segfaults

a lot of fixes, vectorization temp. disabled

fixed trilinear shift size, max error dropped from 19 to 10

fixed several bugs (255 vs 256, signed vs unsigned, bIdx)

minor changes

packed: address arithmetics fixed

shorter code

experiments with pure integer calculations

Lab2RGB max error decreased to 2; need to clean the code

ready for vectorization; need cleaning

vectorized, to be debugged

precision fixed, max error is 2

Lab->XYZ shortened

minor fixes

Lab2RGB_f version fixed, to be completely rewritten using _b code

RGB2Lab_f vectorized

minors

moved to separate file

refactored Lab2RGB to float and int versions

minor fix

Lab2RGB_f vectorized

minor refactoring

Lab2RGBint refactored: process methods, vectorize by 4 pix

Lab2RGB_f int version is done

cleanup extra code

code copied to color.cpp

fixed blue idx bug

optimizations enabled when testing; mulFracConst introduced

divConst -> mulFracConst

calc min time in perf instead of avg

minors

process() slightly sped up

Lab2RGB_f: disabled int version

reinterpret added, minor fixes in names

some warnings fixed

changes transferred to color.cpp

RGB2Lab_f code (and trilinear interpolation code) moved to rgb2lab_faster

whitespace

shift negative fixed

more warnings fixed

"constant condition" warnings fixed, little speed up

minor changes

test_photo decolor fixed

changes copied to test_lab.cpp

idx bounds checking in LUT init

several fixes

WIP: softfloat almost integrated

test_lab partially rewritten to SoftFloat

color.cpp rewritten to SoftFloat

test_lab.cpp: accuracy code added

several fixes

RGB2Lab_b testing fixed

splineBuild() rewritten to SoftFloat

accuracy control improved

rounding fixed

Luv <=> RGB: rewritten to SoftFloat

OCL cvtColor Lab and Lut rewritten to SoftFloat

minor fixes

refactored to new SoftFloat interface

round() -> cvRound, etc.

fixed OCL tests

softfloat.cpp: internal functions made static, unused ones removed

meaningful constants

extra lines removed

unused function removed

unfinished work

it works, need to fix TODOs

refactoring; more calls rewritten

mulFracConst removed

constants made bit exact; minors

changes moved to color.cpp

fixed 1 bug and 4 warnings

OCL: fixed constants

pow(x, _1_3f) replaced by cubeRoot(x)

fixed compilation on MSVC32

magic constants explained

file with internal accuracy&speed tests moved to lab_tetra branch

7 years agofixed 'OpenCVDetectCXXCompiler.cmake' to detect C++11 feature
atinfinity [Sun, 16 Jul 2017 15:20:14 +0000 (00:20 +0900)]
fixed 'OpenCVDetectCXXCompiler.cmake' to detect C++11 feature

7 years agoMerge pull request #9133 from sovrasov:mser_add_test
Alexander Alekhin [Fri, 14 Jul 2017 19:38:05 +0000 (19:38 +0000)]
Merge pull request #9133 from sovrasov:mser_add_test

7 years agoMerge pull request #9086 from catree:improve_solvePnPRansac
Alexander Alekhin [Fri, 14 Jul 2017 17:28:47 +0000 (17:28 +0000)]
Merge pull request #9086 from catree:improve_solvePnPRansac

7 years agoMerge pull request #9091 from alalek:update_ffmpeg
Alexander Alekhin [Fri, 14 Jul 2017 17:25:35 +0000 (17:25 +0000)]
Merge pull request #9091 from alalek:update_ffmpeg

7 years agoMerge pull request #9095 from alalek:fix_gstreamer
Alexander Alekhin [Fri, 14 Jul 2017 17:24:14 +0000 (17:24 +0000)]
Merge pull request #9095 from alalek:fix_gstreamer

7 years agoMerge pull request #9159 from alalek:cmake_target_rename
Alexander Alekhin [Fri, 14 Jul 2017 17:19:25 +0000 (17:19 +0000)]
Merge pull request #9159 from alalek:cmake_target_rename

7 years agoMerge pull request #9157 from zhmu:opencv-dshow-allow-enable-disable-of-debug-messages
Alexander Alekhin [Fri, 14 Jul 2017 17:18:31 +0000 (17:18 +0000)]
Merge pull request #9157 from zhmu:opencv-dshow-allow-enable-disable-of-debug-messages

7 years agoMerge pull request #9130 from alalek:android_define
Alexander Alekhin [Fri, 14 Jul 2017 17:17:24 +0000 (17:17 +0000)]
Merge pull request #9130 from alalek:android_define

7 years agoMerge pull request #9034 from sovrasov:mats_from_initializer_list
Vladislav Sovrasov [Fri, 14 Jul 2017 17:17:09 +0000 (20:17 +0300)]
Merge pull request #9034 from sovrasov:mats_from_initializer_list

Add constructors taking initializer_list for some of OpenCV data types (#9034)

* Add a constructor taking initializer_list for Matx

* Add a constructor taking initializer list for Mat and Mat_

* Add one more method to initialize Mat to the corresponding tutorial

* Add a note how to initialize Matx

* CV_CXX_11->CV_CXX11

7 years agoMerge pull request #8975 from sovrasov:fs_additional_errors
Alexander Alekhin [Fri, 14 Jul 2017 17:13:50 +0000 (17:13 +0000)]
Merge pull request #8975 from sovrasov:fs_additional_errors

7 years agoMerge pull request #9048 from sovrasov:morph_hitmiss_fix
Alexander Alekhin [Fri, 14 Jul 2017 17:13:06 +0000 (17:13 +0000)]
Merge pull request #9048 from sovrasov:morph_hitmiss_fix

imgproc: fix MORPH_HITMISS operation when kernel has no negative values

7 years agoMerge pull request #8914 from stnk20:gstreamer_yuv
Satoshi Tanaka [Fri, 14 Jul 2017 17:11:30 +0000 (02:11 +0900)]
Merge pull request #8914 from stnk20:gstreamer_yuv

Add gstreamer capture capability for some YUV formats (#8914)

* Add gstreamer capture capability for some YUV formats.(only for gstreamer-1.0)

* avoid cross initialization error

* add checking if pipeline is manualpipeline, for compatibility.

7 years agomser: add a couple of new regression tests
Vladislav Sovrasov [Mon, 10 Jul 2017 14:11:04 +0000 (17:11 +0300)]
mser: add a couple of new regression tests

7 years agoMerge pull request #8910 from gylns:mser
gylns [Fri, 14 Jul 2017 16:58:56 +0000 (11:58 -0500)]
Merge pull request #8910 from gylns:mser

fix the MSER history's size issue (#8910)

* simplify growHistory and merge

* add assertion for history's size

* MSER: fix merging components' history

7 years agoMerge pull request #9078 from arrybn:resnet_squeezenet_tests
Alexander Alekhin [Fri, 14 Jul 2017 16:50:25 +0000 (16:50 +0000)]
Merge pull request #9078 from arrybn:resnet_squeezenet_tests

Added tests for ResNet-50 and SqueezeNet v1.1

7 years agoMerge pull request #9074 from alalek:cpu_dispatch_core_hamming
Alexander Alekhin [Fri, 14 Jul 2017 16:48:07 +0000 (16:48 +0000)]
Merge pull request #9074 from alalek:cpu_dispatch_core_hamming

cpu dispatch(core): hamming

7 years agoMerge pull request #9041 from terfendail:filter_avx
Alexander Alekhin [Fri, 14 Jul 2017 16:45:27 +0000 (16:45 +0000)]
Merge pull request #9041 from terfendail:filter_avx

AVX optimized implementation of separable filters migrated

7 years agoMerge pull request #9061 from terfendail:convert_avx
Alexander Alekhin [Fri, 14 Jul 2017 16:43:54 +0000 (16:43 +0000)]
Merge pull request #9061 from terfendail:convert_avx

AVX and SSE4.1 optimized conversion migrated

7 years agoMerge pull request #9082 from terfendail:imgwarp_avx
Alexander Alekhin [Fri, 14 Jul 2017 16:42:42 +0000 (16:42 +0000)]
Merge pull request #9082 from terfendail:imgwarp_avx

AVX and SSE4.1 optimized implementation of resize and warp functions migrated

7 years agoMerge pull request #9088 from sovrasov:no_nostl
Alexander Alekhin [Fri, 14 Jul 2017 16:26:03 +0000 (16:26 +0000)]
Merge pull request #9088 from sovrasov:no_nostl

core: get rid of OPENCV_NOSTL definition

7 years agoMerge pull request #9090 from vpisarev:dnn_optim_scale_concat
Alexander Alekhin [Fri, 14 Jul 2017 16:21:24 +0000 (16:21 +0000)]
Merge pull request #9090 from vpisarev:dnn_optim_scale_concat

7 years agoMerge pull request #9127 from willbrazil:master
Alexander Alekhin [Fri, 14 Jul 2017 16:12:48 +0000 (16:12 +0000)]
Merge pull request #9127 from willbrazil:master

Fix typo in harris corner detection tutorial.

7 years agoMerge pull request #9098 from savuor:fix/luv_div
Alexander Alekhin [Fri, 14 Jul 2017 15:46:02 +0000 (15:46 +0000)]
Merge pull request #9098 from savuor:fix/luv_div

7 years agooptimize out scaleLayer & concatLayer whenever possible
Vadim Pisarevsky [Tue, 4 Jul 2017 14:23:47 +0000 (17:23 +0300)]
optimize out scaleLayer & concatLayer whenever possible

fixed problem in concat layer by disabling memory re-use in layers with multiple inputs

trying to fix the tests when Halide is used to run deep nets

another attempt to fix Halide tests

see if the Halide tests will pass with concat layer fusion turned off

trying to fix failures in halide tests; another try

one more experiment to make halide_concat & halide_enet tests pass

continue attempts to fix halide tests

moving on

uncomment parallel concat layer

seemingly fixed failures in Halide tests and re-enabled concat layer fusion; thanks to dkurt for the patch

7 years agoMerge pull request #9042 from terfendail:haar_avx
Alexander Alekhin [Fri, 14 Jul 2017 15:05:11 +0000 (15:05 +0000)]
Merge pull request #9042 from terfendail:haar_avx

AVX optimized implementation of haar migrated to separate file

7 years agoMerge pull request #9021 from terfendail:corner_avx
Alexander Alekhin [Fri, 14 Jul 2017 14:58:06 +0000 (14:58 +0000)]
Merge pull request #9021 from terfendail:corner_avx

7 years agoMerge pull request #9027 from terfendail:undistort_avx
Alexander Alekhin [Fri, 14 Jul 2017 14:56:42 +0000 (14:56 +0000)]
Merge pull request #9027 from terfendail:undistort_avx

7 years agoMerge pull request #9063 from alalek:hamming_perf_test
Alexander Alekhin [Fri, 14 Jul 2017 14:46:16 +0000 (14:46 +0000)]
Merge pull request #9063 from alalek:hamming_perf_test

7 years agoMerge pull request #9093 from wzw-intel:histogram
Alexander Alekhin [Fri, 14 Jul 2017 14:38:55 +0000 (14:38 +0000)]
Merge pull request #9093 from wzw-intel:histogram

7 years agoMerge pull request #9122 from ivsgroup:fix_msvc_virtual_destructor
Alexander Alekhin [Fri, 14 Jul 2017 14:37:22 +0000 (14:37 +0000)]
Merge pull request #9122 from ivsgroup:fix_msvc_virtual_destructor

7 years agoMerge pull request #9110 from patrikhuber:fix-msvc1911-cmake
Alexander Alekhin [Fri, 14 Jul 2017 14:37:01 +0000 (14:37 +0000)]
Merge pull request #9110 from patrikhuber:fix-msvc1911-cmake

7 years agoMerge pull request #9153 from sovrasov:fix_tangent_dist_flag
Alexander Alekhin [Fri, 14 Jul 2017 14:35:51 +0000 (14:35 +0000)]
Merge pull request #9153 from sovrasov:fix_tangent_dist_flag

7 years agoMerge pull request #9094 from ArkadiuszRaj:fix-aravis-dependency
Alexander Alekhin [Fri, 14 Jul 2017 14:31:57 +0000 (14:31 +0000)]
Merge pull request #9094 from ArkadiuszRaj:fix-aravis-dependency

7 years agoMerge pull request #9024 from tomoaki0705:featureDispatchAccumulate
Alexander Alekhin [Fri, 14 Jul 2017 14:30:05 +0000 (14:30 +0000)]
Merge pull request #9024 from tomoaki0705:featureDispatchAccumulate

7 years agoMerge pull request #9058 from alalek:dnn_minor_fixes
Alexander Alekhin [Fri, 14 Jul 2017 09:53:09 +0000 (09:53 +0000)]
Merge pull request #9058 from alalek:dnn_minor_fixes

7 years agobuild: added DEBUG build guard
Alexander Alekhin [Thu, 13 Jul 2017 21:56:09 +0000 (00:56 +0300)]
build: added DEBUG build guard

To prevent linkage of binary incompatible DEBUG/RELEASE binaries/runtimes

7 years agocmake: rename cpufeatures target
Alexander Alekhin [Thu, 13 Jul 2017 16:23:04 +0000 (19:23 +0300)]
cmake: rename cpufeatures target

It is configurable via OPENCV_CPUFEATURES_TARGET_NAME variable

7 years agodnn: cleanup dispatched code, fix SIMD128 types
Alexander Alekhin [Thu, 13 Jul 2017 15:42:36 +0000 (18:42 +0300)]
dnn: cleanup dispatched code, fix SIMD128 types

7 years agodnn: code cleanup, refactor detection output layer
Alexander Alekhin [Thu, 13 Jul 2017 13:34:23 +0000 (16:34 +0300)]
dnn: code cleanup, refactor detection output layer

7 years agodnn: some minor fixes in docs, indentation, unused code
Alexander Alekhin [Fri, 30 Jun 2017 15:46:00 +0000 (18:46 +0300)]
dnn: some minor fixes in docs, indentation, unused code

7 years agoMerge pull request #9111 from vpisarev:dnn_optim_avx1
Alexander Alekhin [Thu, 13 Jul 2017 12:27:05 +0000 (12:27 +0000)]
Merge pull request #9111 from vpisarev:dnn_optim_avx1

7 years agodshow: Only show debugging messages if environment variable OPENCV_DSHOW_DEBUG is...
Rink Springer [Thu, 13 Jul 2017 09:44:20 +0000 (11:44 +0200)]
dshow: Only show debugging messages if environment variable OPENCV_DSHOW_DEBUG is explicitly set to non-zero

Based on discussion at: https://github.com/opencv/opencv/pull/9051

7 years agocalib3d: fix not working CALIB_FIX_TANGENT_DIST flag
Vladislav Sovrasov [Thu, 13 Jul 2017 07:49:35 +0000 (10:49 +0300)]
calib3d: fix not working CALIB_FIX_TANGENT_DIST flag

7 years agoMerge pull request #9107 from alalek:static_analyze_python
Alexander Alekhin [Wed, 12 Jul 2017 16:45:13 +0000 (16:45 +0000)]
Merge pull request #9107 from alalek:static_analyze_python

7 years agoMerge pull request #9142 from alalek:vzeroupper_guard_unused_warning
Alexander Alekhin [Wed, 12 Jul 2017 16:44:00 +0000 (16:44 +0000)]
Merge pull request #9142 from alalek:vzeroupper_guard_unused_warning

7 years agoMerge pull request #9149 from alalek:ocl_fp16_test
Alexander Alekhin [Wed, 12 Jul 2017 16:23:33 +0000 (16:23 +0000)]
Merge pull request #9149 from alalek:ocl_fp16_test

7 years agocore(test): fix input data for OCL FP16 test
Alexander Alekhin [Wed, 12 Jul 2017 15:51:11 +0000 (18:51 +0300)]
core(test): fix input data for OCL FP16 test

7 years agoMerge pull request #9145 from alalek:ipp_minmaxidx_nan
Alexander Alekhin [Wed, 12 Jul 2017 15:37:53 +0000 (15:37 +0000)]
Merge pull request #9145 from alalek:ipp_minmaxidx_nan

7 years agoipp(minmaxIdx): disable SSE4.2 optimizations for 32f datatype
Alexander Alekhin [Wed, 12 Jul 2017 13:06:18 +0000 (16:06 +0300)]
ipp(minmaxIdx): disable SSE4.2 optimizations for 32f datatype

NaN values handling issue

7 years agoadd std::string overload for cv::read()
PkLab.net [Fri, 26 May 2017 17:14:58 +0000 (19:14 +0200)]
add std::string overload for cv::read()

7 years agocore: forbid conversion real->int in some cases in FileStorage
Vladislav Sovrasov [Fri, 23 Jun 2017 14:03:13 +0000 (17:03 +0300)]
core: forbid conversion real->int in some cases in FileStorage

7 years agoFix wrong chessboard generation in gen_pattern tool
Vladislav Sovrasov [Tue, 11 Jul 2017 14:39:52 +0000 (17:39 +0300)]
Fix wrong chessboard generation in gen_pattern tool

7 years agobuild: fix unused variable warning for vzeroupper guard
Alexander Alekhin [Tue, 11 Jul 2017 13:46:35 +0000 (16:46 +0300)]
build: fix unused variable warning for vzeroupper guard

7 years agoMerge pull request #9131 from dkurt:fix_eltwise_layer
Alexander Alekhin [Tue, 11 Jul 2017 07:18:41 +0000 (07:18 +0000)]
Merge pull request #9131 from dkurt:fix_eltwise_layer

7 years agoMerge pull request #9134 from mshabunin:fix-static-6
Alexander Alekhin [Mon, 10 Jul 2017 20:49:36 +0000 (20:49 +0000)]
Merge pull request #9134 from mshabunin:fix-static-6

7 years agoFixed several issues found by static analysis (Windows specific)
Maksim Shabunin [Mon, 10 Jul 2017 20:14:02 +0000 (23:14 +0300)]
Fixed several issues found by static analysis (Windows specific)

7 years agoEltwise layer fixes
dkurt [Mon, 10 Jul 2017 09:58:11 +0000 (12:58 +0300)]
Eltwise layer fixes

7 years agobuild: detect Android via '__ANDROID__' macro
Alexander Alekhin [Mon, 10 Jul 2017 09:43:59 +0000 (12:43 +0300)]
build: detect Android via '__ANDROID__' macro

https://sourceforge.net/p/predef/wiki/OperatingSystems

7 years agoFix typo: 'right direction' to 'y direction'.
Guilherme William Guedes [Sun, 9 Jul 2017 22:57:55 +0000 (15:57 -0700)]
Fix typo: 'right direction' to 'y direction'.

Fix typo: 'right direction' to 'y direction' in harris corner detector tutorial.

7 years agocore/bufferpool.hpp: let msvc accept a non virtual protected destructor
Pascal Thomet [Fri, 7 Jul 2017 19:04:12 +0000 (21:04 +0200)]
core/bufferpool.hpp: let msvc accept a non virtual protected destructor

BufferPoolController has a non virtual protected destructor (which is legitimate)

However, Visual Studio sees this as a bug, if you enable more warnings, like below
```
add_compile_options(/W3)     # level 3 warnings
add_compile_options(/we4265) # warning about missing virtual destructors
```

This is a proposition in order to silence this warning.

See https://github.com/ivsgroup/boost_warnings_minimal_demo for a demo of the same problem
with boost/exception.hpp

7 years agoMerge pull request #9115 from alalek:ocl_async_cleanup_queue
Maksim Shabunin [Fri, 7 Jul 2017 12:05:34 +0000 (12:05 +0000)]
Merge pull request #9115 from alalek:ocl_async_cleanup_queue

7 years agoocl: async cl_buffer cleanup queue (for event callback)
Alexander Alekhin [Thu, 6 Jul 2017 14:57:05 +0000 (17:57 +0300)]
ocl: async cl_buffer cleanup queue (for event callback)

7 years agodispatch accumulate series
Tomoaki Teshima [Fri, 7 Jul 2017 09:45:30 +0000 (18:45 +0900)]
dispatch accumulate series
 - use universal intrinsic for base
 - dispatch for float/double version using AVX
 - AVX2 optimization not done yet

7 years agoMerge pull request #9105 from alalek:ocl_update_event_callback
Maksim Shabunin [Fri, 7 Jul 2017 08:23:46 +0000 (08:23 +0000)]
Merge pull request #9105 from alalek:ocl_update_event_callback

7 years agoreuse AVX2-optimized kernels for AVX1 CPUs (like IvyBridge)
Vadim Pisarevsky [Thu, 6 Jul 2017 18:36:59 +0000 (21:36 +0300)]
reuse AVX2-optimized kernels for AVX1 CPUs (like IvyBridge)

7 years agoAdded detection of MSVC1911
Patrik Huber [Thu, 6 Jul 2017 18:02:19 +0000 (19:02 +0100)]
Added detection of MSVC1911

This is VS2017 Preview. It makes sure the version is detected properly and the INSTALL target then correctly installs to x64/vc15 (the same as MSVC1910).

7 years agostatic analyze: python dynamic_cast checks
Alexander Alekhin [Thu, 6 Jul 2017 12:57:12 +0000 (15:57 +0300)]
static analyze: python dynamic_cast checks

7 years agoocl: rework events handling with clSetEventCallback
Alexander Alekhin [Thu, 6 Jul 2017 10:25:32 +0000 (13:25 +0300)]
ocl: rework events handling with clSetEventCallback

7 years agoAVX optimized implementation of undistort migrated to separate file
Vitaly Tuzov [Wed, 28 Jun 2017 16:57:26 +0000 (19:57 +0300)]
AVX optimized implementation of undistort migrated to separate file

7 years agomagic constants explained
Rostislav Vasilikhin [Wed, 5 Jul 2017 21:23:58 +0000 (00:23 +0300)]
magic constants explained

7 years agoOCL code fixed, fix for NEON added
Rostislav Vasilikhin [Wed, 5 Jul 2017 18:52:13 +0000 (21:52 +0300)]
OCL code fixed, fix for NEON added

7 years agoRGB2Luv_f: R, G, B limited to [0, 1]
Rostislav Vasilikhin [Wed, 5 Jul 2017 18:06:14 +0000 (21:06 +0300)]
RGB2Luv_f: R, G, B limited to [0, 1]

7 years agoLuv: singularities fixed
Rostislav Vasilikhin [Wed, 5 Jul 2017 17:36:30 +0000 (20:36 +0300)]
Luv: singularities fixed