platform/upstream/opencv.git
4 years agoMerge pull request #16409 from smirnov-alexey:as/gapi_fix_standalone_streaming
Alexander Alekhin [Fri, 24 Jan 2020 08:47:57 +0000 (08:47 +0000)]
Merge pull request #16409 from smirnov-alexey:as/gapi_fix_standalone_streaming

4 years agoFix linkage in standalone mode
Smirnov Alexey [Wed, 22 Jan 2020 14:51:16 +0000 (17:51 +0300)]
Fix linkage in standalone mode

4 years agoMerge remote-tracking branch 'upstream/3.4' into merge-3.4
Alexander Alekhin [Wed, 22 Jan 2020 13:58:30 +0000 (16:58 +0300)]
Merge remote-tracking branch 'upstream/3.4' into merge-3.4

4 years agoMerge pull request #16408 from alalek:dnn_test_tolerance_myriad
Alexander Alekhin [Wed, 22 Jan 2020 13:56:15 +0000 (13:56 +0000)]
Merge pull request #16408 from alalek:dnn_test_tolerance_myriad

4 years agoMerge pull request #16146 from pmur:reg_16137x2
Alexander Alekhin [Wed, 22 Jan 2020 13:53:51 +0000 (13:53 +0000)]
Merge pull request #16146 from pmur:reg_16137x2

4 years agodnn(test): adjust test tolerance for MYRIAD
Alexander Alekhin [Wed, 22 Jan 2020 11:42:52 +0000 (14:42 +0300)]
dnn(test): adjust test tolerance for MYRIAD

4 years agoMerge pull request #16358 from ashishkrshrivastava:opencv-2
Alexander Alekhin [Wed, 22 Jan 2020 09:57:41 +0000 (09:57 +0000)]
Merge pull request #16358 from ashishkrshrivastava:opencv-2

4 years agoMerge pull request #16400 from alalek:issue_7001
Alexander Alekhin [Tue, 21 Jan 2020 19:16:39 +0000 (19:16 +0000)]
Merge pull request #16400 from alalek:issue_7001

4 years agoMerge pull request #16393 from alalek:build_videoio_fix_15020
Alexander Alekhin [Tue, 21 Jan 2020 19:16:00 +0000 (19:16 +0000)]
Merge pull request #16393 from alalek:build_videoio_fix_15020

4 years agocmake: apply '-Wl,--as-needed' linker option
Alexander Alekhin [Tue, 21 Jan 2020 14:19:30 +0000 (17:19 +0300)]
cmake: apply '-Wl,--as-needed' linker option

4 years agoMerge pull request #16391 from l-bat:ngraph_lrn
Alexander Alekhin [Tue, 21 Jan 2020 10:08:54 +0000 (10:08 +0000)]
Merge pull request #16391 from l-bat:ngraph_lrn

4 years agoSupport lrn with SPATIAL_NRM
Liubov Batanina [Mon, 20 Jan 2020 14:07:35 +0000 (17:07 +0300)]
Support lrn with SPATIAL_NRM

4 years agoMerge pull request #16389 from alalek:core_fix_msa_build
Alexander Alekhin [Mon, 20 Jan 2020 16:31:38 +0000 (16:31 +0000)]
Merge pull request #16389 from alalek:core_fix_msa_build

4 years agobuild: workaround for FORTIFY_SOURCE=2 builds on Alpine
Alexander Alekhin [Mon, 20 Jan 2020 14:45:56 +0000 (17:45 +0300)]
build: workaround for FORTIFY_SOURCE=2 builds on Alpine

4 years agocore(SIMD): fix MSA build - add v_reduce_min/max for u8/s8
Alexander Alekhin [Mon, 20 Jan 2020 12:08:51 +0000 (15:08 +0300)]
core(SIMD): fix MSA build - add v_reduce_min/max for u8/s8

4 years agoMerge pull request #16309 from ganesh-k13:bugfix/imdecode-resize
Ganesh Kathiresan [Mon, 20 Jan 2020 11:00:45 +0000 (16:30 +0530)]
Merge pull request #16309 from ganesh-k13:bugfix/imdecode-resize

* Added flag handlers for imread and imdecode | Issue 16203

Undo imread change

Added Imread resize tests

Added imdecode flags check

Added imdecode tests for resize

Removed trailing whitespace

Removed IMREAD_IGNORE_ORIENTATION check

Added else condition

Removed IMREAD_IGNORE_ORIENTATION check in decode

Added HAVE_JPEG guards

Added static_cast for Win32

Added resize for non jpegs

Added tests for non jpeg resize case

Fixed resize value in assert

Changed tests to Value-Parameterized Tests

Changed tests to Value-Parameterized Tests | handled >> in cpp

Changed tests to Value-Parameterized Tests | removed trailing whitespace

* imgcodecs: update test

4 years agoMerge pull request #16385 from alalek:ts_update_optional_message
Alexander Alekhin [Mon, 20 Jan 2020 10:54:40 +0000 (10:54 +0000)]
Merge pull request #16385 from alalek:ts_update_optional_message

4 years agoMerge pull request #16378 from alalek:dnn_getMemoryShapes_error_dump_blobs
Alexander Alekhin [Mon, 20 Jan 2020 10:49:17 +0000 (10:49 +0000)]
Merge pull request #16378 from alalek:dnn_getMemoryShapes_error_dump_blobs

4 years agoMerge pull request #16377 from alalek:fix_16373
Alexander Alekhin [Mon, 20 Jan 2020 10:48:21 +0000 (10:48 +0000)]
Merge pull request #16377 from alalek:fix_16373

4 years agoConvolutional padding parsing modified
ashishiva3@gmail.com [Wed, 15 Jan 2020 13:31:01 +0000 (19:01 +0530)]
Convolutional padding parsing modified

4 years agots: move message about misused 'optional' under the configuration parameter
Alexander Alekhin [Sun, 19 Jan 2020 16:09:40 +0000 (16:09 +0000)]
ts: move message about misused 'optional' under the configuration parameter

4 years agoMerge pull request #15929 from sturkmen72:patch-3
Alexander Alekhin [Sun, 19 Jan 2020 12:45:28 +0000 (12:45 +0000)]
Merge pull request #15929 from sturkmen72:patch-3

4 years agoMerge pull request #16372 from alalek:backport_16350
Alexander Alekhin [Sat, 18 Jan 2020 17:57:46 +0000 (17:57 +0000)]
Merge pull request #16372 from alalek:backport_16350

4 years agoMerge pull request #16350 from hannesa2:fixLintIssuesAndroid
Alexander Alekhin [Sat, 18 Jan 2020 17:57:02 +0000 (17:57 +0000)]
Merge pull request #16350 from hannesa2:fixLintIssuesAndroid

4 years agoMerge pull request #15488 from ChipKerchner:vectorizeMinMax2
Chip Kerchner [Fri, 17 Jan 2020 16:37:35 +0000 (11:37 -0500)]
Merge pull request #15488 from ChipKerchner:vectorizeMinMax2

Vectorize minMaxIdx functions

* Updated documentation and intrinsic tests for v_reduce

* Add other files back in from the forced push

* Prevent an constant overflow with v_reduce for int8 type

* Another alternative to fix constant overflow warning.

* Fix another compiler warning.

* Update comments and change comparison form to be consistent with other vectorized loops.

* Change return type of v_reduce_min & max for v_uint8 and v_uint16 to be same as lane type.

* Cast v_reduce functions to int to avoid overflow. Reduce number of parameters in MINMAXIDX_REDUCE macro.

* Restore cast type for v_reduce_min & max to LaneType

4 years agoMerge pull request #16273 from JulienMaille:wrapper_available_target
Julien [Fri, 17 Jan 2020 16:24:37 +0000 (17:24 +0100)]
Merge pull request #16273 from JulienMaille:wrapper_available_target

* add a wrapper for getAvailableTargets

* add java wrapper on Target enum

4 years agodnn(onnx): handle unaligned access in ONNX importer
Alexander Alekhin [Fri, 17 Jan 2020 13:38:58 +0000 (16:38 +0300)]
dnn(onnx): handle unaligned access in ONNX importer

4 years agoMerge pull request #16376 from JulienMaille:cuda-dnn-for-older-gpus
Alexander Alekhin [Fri, 17 Jan 2020 15:42:50 +0000 (15:42 +0000)]
Merge pull request #16376 from JulienMaille:cuda-dnn-for-older-gpus

4 years agoFIX: disable dnn cuda input_shortcut on _half for CC<5.3
Julien Maille [Fri, 17 Jan 2020 13:21:25 +0000 (14:21 +0100)]
FIX: disable dnn cuda input_shortcut on _half for CC<5.3

4 years agofix lint errors and warnings ins Android samples
Hannes Achleitner [Wed, 15 Jan 2020 05:29:49 +0000 (06:29 +0100)]
fix lint errors and warnings ins Android samples

backporting of commit 4057e81b7649d8d814908157478456c6889271cd

4 years agoMerge pull request #16367 from dkurt:dnn_ie_ngraph_mvn
Alexander Alekhin [Fri, 17 Jan 2020 09:12:36 +0000 (09:12 +0000)]
Merge pull request #16367 from dkurt:dnn_ie_ngraph_mvn

4 years agoMerge pull request #16369 from kraj:gcc10
Alexander Alekhin [Fri, 17 Jan 2020 08:58:19 +0000 (08:58 +0000)]
Merge pull request #16369 from kraj:gcc10

4 years agoMerge pull request #16294 from h6197627:master
Alexander Alekhin [Fri, 17 Jan 2020 08:53:09 +0000 (08:53 +0000)]
Merge pull request #16294 from h6197627:master

4 years agocmake: fix Linux system OpenEXR detection
h6197627 [Mon, 6 Jan 2020 13:52:40 +0000 (15:52 +0200)]
cmake: fix Linux system OpenEXR detection

4 years agocarotene: Replace ipcp-unit-growth with ipa-cp-unit-growth on gcc >= 10
Khem Raj [Thu, 16 Jan 2020 16:52:00 +0000 (08:52 -0800)]
carotene: Replace ipcp-unit-growth with ipa-cp-unit-growth on gcc >= 10

gcc 10+ has renamed this option, therefore check for gcc version before
deciding which name to use for opt parameter

Signed-off-by: Khem Raj <raj.khem@gmail.com>
4 years agoMerge pull request #16063 from YashasSamaga:cuda4dnn-shortcut-unequal
Yashas Samaga B L [Thu, 16 Jan 2020 18:54:00 +0000 (00:24 +0530)]
Merge pull request #16063 from YashasSamaga:cuda4dnn-shortcut-unequal

support eltwise sum with different number of input channels in CUDA backend

* add shortcut primitive

* add offsets in shortcut kernel

* skip tests involving more than two inputs

* remove redundant modulus operation

* support multiple inputs

* remove whole file indentation

* skip acc in0 trunc test if weighted

* use shortcut iff channels are unequal

4 years agoMerge pull request #16349 from hannesa2:RemoveEclipseRelicts
Alexander Alekhin [Thu, 16 Jan 2020 16:53:06 +0000 (16:53 +0000)]
Merge pull request #16349 from hannesa2:RemoveEclipseRelicts

4 years agoMerge pull request #16325 from alalek:core_simd_neon_fix_alignment_lut
Alexander Alekhin [Thu, 16 Jan 2020 15:18:56 +0000 (15:18 +0000)]
Merge pull request #16325 from alalek:core_simd_neon_fix_alignment_lut

4 years agoMerge pull request #16351 from hannesa2:finishAndroidVersionGradleStyle
Alexander Alekhin [Thu, 16 Jan 2020 15:17:17 +0000 (15:17 +0000)]
Merge pull request #16351 from hannesa2:finishAndroidVersionGradleStyle

4 years agoMVN support through nGraph
Dmitry Kurtaev [Thu, 16 Jan 2020 15:04:10 +0000 (18:04 +0300)]
MVN support through nGraph

4 years agoMerge pull request #16257 from ianare:more-pythonic
Alexander Alekhin [Thu, 16 Jan 2020 13:57:08 +0000 (13:57 +0000)]
Merge pull request #16257 from ianare:more-pythonic

4 years agodnn: dump layer's blobs on getMemoryShapes() error
Alexander Alekhin [Thu, 16 Jan 2020 11:31:43 +0000 (14:31 +0300)]
dnn: dump layer's blobs on getMemoryShapes() error

4 years agoMerge pull request #16363 from JulienMaille:cuda-dnn-for-older-gpus
Alexander Alekhin [Thu, 16 Jan 2020 11:08:04 +0000 (11:08 +0000)]
Merge pull request #16363 from JulienMaille:cuda-dnn-for-older-gpus

4 years agoFix: rsqrt(float) was improperly put in the ifdef for half
Julien [Wed, 15 Jan 2020 20:06:25 +0000 (21:06 +0100)]
Fix: rsqrt(float) was improperly put in the ifdef for half

4 years agofix lint errors and warnings ins Android samples
Hannes Achleitner [Wed, 15 Jan 2020 05:29:49 +0000 (06:29 +0100)]
fix lint errors and warnings ins Android samples

4 years agoMerge pull request #16361 from alalek:android_fix_javadoc
Alexander Alekhin [Wed, 15 Jan 2020 19:42:35 +0000 (19:42 +0000)]
Merge pull request #16361 from alalek:android_fix_javadoc

4 years agoMerge pull request #16357 from dkurt:dnn_ie_reset_myriad
Alexander Alekhin [Wed, 15 Jan 2020 17:03:09 +0000 (17:03 +0000)]
Merge pull request #16357 from dkurt:dnn_ie_reset_myriad

4 years agoandroid: fix javadoc
Alexander Alekhin [Wed, 15 Jan 2020 16:39:32 +0000 (19:39 +0300)]
android: fix javadoc

4 years agoDifferent way to reset Myriad device
Dmitry Kurtaev [Wed, 15 Jan 2020 13:22:00 +0000 (16:22 +0300)]
Different way to reset Myriad device

4 years agoMerge pull request #16354 from dkurt:dnn_ie_custom_cpu_layers
Alexander Alekhin [Wed, 15 Jan 2020 15:33:18 +0000 (15:33 +0000)]
Merge pull request #16354 from dkurt:dnn_ie_custom_cpu_layers

4 years agoMerge pull request #16218 from JulienMaille:cuda-dnn-for-older-gpus
Julien [Wed, 15 Jan 2020 15:28:37 +0000 (16:28 +0100)]
Merge pull request #16218 from JulienMaille:cuda-dnn-for-older-gpus

Enable cuda4dnn on hardware without support for __half

* Enable cuda4dnn on hardware without support for half (ie. compute capability < 5.3)

Update CMakeLists.txt

Lowered minimum CC to 3.0

* UPD: added ifdef on new copy kernel

* added fp16 support detection at runtime

* Clarified #if condition on atomicAdd definition

* More explicit CMake error message

4 years agoMerge pull request #16353 from dkurt:uninitialized_value
Alexander Alekhin [Wed, 15 Jan 2020 14:49:45 +0000 (14:49 +0000)]
Merge pull request #16353 from dkurt:uninitialized_value

4 years agoWrap custom OpenCV layers to try-catch
Dmitry Kurtaev [Wed, 15 Jan 2020 12:10:17 +0000 (15:10 +0300)]
Wrap custom OpenCV layers to try-catch

4 years agoFix uninitialized value
Dmitry Kurtaev [Wed, 15 Jan 2020 09:58:12 +0000 (12:58 +0300)]
Fix uninitialized value

4 years agoremove Eclipse relicts in Android sample
Hannes Achleitner [Wed, 15 Jan 2020 06:19:58 +0000 (07:19 +0100)]
remove Eclipse relicts in Android sample

4 years agofinish Android version info to Gradle style
Hannes Achleitner [Wed, 15 Jan 2020 05:34:42 +0000 (06:34 +0100)]
finish Android version info to Gradle style

4 years agoMerge remote-tracking branch 'upstream/3.4' into merge-3.4
Alexander Alekhin [Tue, 14 Jan 2020 14:04:22 +0000 (17:04 +0300)]
Merge remote-tracking branch 'upstream/3.4' into merge-3.4

4 years agoMerge pull request #16291 from dkurt:dnn_onnx_graph_simplifier
Alexander Alekhin [Tue, 14 Jan 2020 12:45:59 +0000 (12:45 +0000)]
Merge pull request #16291 from dkurt:dnn_onnx_graph_simplifier

4 years agoMerge pull request #16199 from sajarindider:computeGradientX
Alexander Alekhin [Tue, 14 Jan 2020 12:42:28 +0000 (12:42 +0000)]
Merge pull request #16199 from sajarindider:computeGradientX

4 years agoMerge pull request #16344 from l-bat:ngraph_normalize
Alexander Alekhin [Tue, 14 Jan 2020 11:55:37 +0000 (11:55 +0000)]
Merge pull request #16344 from l-bat:ngraph_normalize

4 years agoFixed small inefficiency in seamless clone init
sajarindider [Thu, 19 Dec 2019 05:46:46 +0000 (00:46 -0500)]
Fixed small inefficiency in seamless clone init

4 years agoEnable acrossSpatial normalizeL2 on Myriad
Liubov Batanina [Tue, 14 Jan 2020 09:51:19 +0000 (12:51 +0300)]
Enable acrossSpatial normalizeL2 on Myriad

4 years agoONNX graphs simplifier
Dmitry Kurtaev [Mon, 6 Jan 2020 11:03:05 +0000 (14:03 +0300)]
ONNX graphs simplifier

4 years agoMerge pull request #16232 from dkurt:dnn_ie_ngraph_fix_myriad_tests
Alexander Alekhin [Mon, 13 Jan 2020 16:59:45 +0000 (16:59 +0000)]
Merge pull request #16232 from dkurt:dnn_ie_ngraph_fix_myriad_tests

4 years agoMerge pull request #15915 from VadimLevin:dev/norm_fix
Vadim Levin [Mon, 13 Jan 2020 15:11:34 +0000 (18:11 +0300)]
Merge pull request #15915 from VadimLevin:dev/norm_fix

Fix implicit conversion from array to scalar in python bindings

* Fix wrong conversion behavior for primitive types

  - Introduce ArgTypeInfo namedtuple instead of plain tuple.
    If strict conversion parameter for type is set to true, it is
    handled like object argument in PyArg_ParseTupleAndKeywords and
    converted to concrete type with the appropriate pyopencv_to function
    call.
  - Remove deadcode and unused variables.
  - Fix implicit conversion from numpy array with 1 element to scalar
  - Fix narrowing conversion to size_t type.

* Fix wrong conversion behavior for primitive types

  - Introduce ArgTypeInfo namedtuple instead of plain tuple.
    If strict conversion parameter for type is set to true, it is
    handled like object argument in PyArg_ParseTupleAndKeywords and
    converted to concrete type with the appropriate pyopencv_to function
    call.
  - Remove deadcode and unused variables.
  - Fix implicit conversion from numpy array with 1 element to scalar
  - Fix narrowing conversion to size_t type.·
  - Enable tests with wrong conversion behavior
  - Restrict passing None as value
  - Restrict bool to integer/floating types conversion

* Add PyIntType support for Python 2

* Remove possible narrowing conversion of size_t

* Bindings conversion update

  - Remove unused macro
  - Add better conversion for types to numpy types descriptors
  - Add argument name to fail messages
  - NoneType treated as a valid argument. Better handling will be added
    as a standalone patch

* Add descriptor specialization for size_t

* Add check for signed to unsigned integer conversion safety

  - If signed integer is positive it can be safely converted
    to unsigned
  - Add check for plain python 2 objects
  - Add check for numpy scalars
  - Add simple type_traits implementation for better code style

* Resolve type "overflow" false negative in safe casting check

 - Move type_traits to separate header

* Add copyright message to type_traits.hpp

* Limit conversion scope for integral numpy types

  - Made canBeSafelyCasted specialized only for size_t, so
    type_traits header became unused and was removed.
  - Added clarification about descriptor pointer

4 years agoDisable some tests for Myriad target of nGraph
Dmitry Kurtaev [Tue, 24 Dec 2019 10:34:33 +0000 (13:34 +0300)]
Disable some tests for Myriad target of nGraph

Add lightweight IE hardware targets checks

nGraph: Concat with paddings

Enable more nGraph tests

Restore FP32->FP16 for GPU plugin of IE

try to fix buildbot

Use lightweight IE targets check only starts from R4

4 years agoMerge pull request #16335 from berak:fix_ml_python_digits_samples_3.4
Alexander Alekhin [Mon, 13 Jan 2020 12:16:40 +0000 (12:16 +0000)]
Merge pull request #16335 from berak:fix_ml_python_digits_samples_3.4

4 years agoMerge pull request #16221 from TolyaTalamanov:at/fix-g_typed_kernel_m
Anatoliy Talamanov [Mon, 13 Jan 2020 11:54:10 +0000 (14:54 +0300)]
Merge pull request #16221 from TolyaTalamanov:at/fix-g_typed_kernel_m

G-API: Fix G_TYPED_KERNEL_M macro

* Fix G_TYPED_KERNEL_M macro

* Fixes

* Fix windows build
* Fix doxygen

* Added several macros

* Add overloads for G_TYPED_KERNEL

4 years agoml/python: fix digits samples(3.4)
berak [Mon, 13 Jan 2020 11:26:28 +0000 (12:26 +0100)]
ml/python: fix digits samples(3.4)

4 years agoMerge remote-tracking branch 'upstream/3.4' into merge-3.4
Alexander Alekhin [Sun, 12 Jan 2020 09:34:34 +0000 (09:34 +0000)]
Merge remote-tracking branch 'upstream/3.4' into merge-3.4

4 years agoMerge pull request #16276 from themechanicalcoder:video-tutorial
Gourav Roy [Sun, 12 Jan 2020 08:34:59 +0000 (14:04 +0530)]
Merge pull request #16276 from themechanicalcoder:video-tutorial

* Added java code for meanshift and optical_flow

* added java code for module video

* added appropriate spaces in codes

* converted absolute path to command line arguments

* added spaces at appropriate places

4 years agoMerge pull request #14666 from ihsan314:file_io_xml_yml
ihsan314 [Sun, 12 Jan 2020 08:32:12 +0000 (00:32 -0800)]
Merge pull request #14666 from ihsan314:file_io_xml_yml

Python code examples for file IO in xml and yml format

* Initial "Pythonization" of file_input_output.cpp

* Moved file_input_output.py to correct location

* Nearly done Pythonizing file_input_output.cpp

* Python equivalent of file_input_output.py created

* Started Pythonizing camera_calibration.cpp

* Completed Python tutorial/sample code for file_input_output

* Resolved whitespace issues

* Removed tabs in file_input_output.cpp

* Patched import order and wrapped code in main function

* Changed string to docstring format in help file

* Updated link to Python example code

4 years agocore(simd): fix NEON alignmnet issue
Alexander Alekhin [Sat, 11 Jan 2020 18:39:50 +0000 (18:39 +0000)]
core(simd): fix NEON alignmnet issue

4 years agoMerge pull request #16270 from dkurt:dnn_sort_tf_text_graphs
Alexander Alekhin [Sat, 11 Jan 2020 17:39:13 +0000 (17:39 +0000)]
Merge pull request #16270 from dkurt:dnn_sort_tf_text_graphs

4 years agoMerge pull request #16236 from alalek:fix_core_simd_emulator
Alexander Alekhin [Fri, 10 Jan 2020 18:31:02 +0000 (21:31 +0300)]
Merge pull request #16236 from alalek:fix_core_simd_emulator

* core: fix intrin_cpp, allow to build modules with SIMD emulator

* core(arithm): fix v_zero initialization

* core(simd): 'strict' types for binary/bitwise operations

* features2d: avoid aligned load issue in GCC 5.4 with emulated SIMD

* core(simd): alignment checks in SIMD emulator

4 years agoMerge pull request #16240 from olramde:olramde
olramde [Fri, 10 Jan 2020 13:18:31 +0000 (22:18 +0900)]
Merge pull request #16240 from olramde:olramde

* Changed plus operator to os.path.join()

* Remove '/' from PATH

4 years agoMerge pull request #16302 from sthibaul:dlopen-glibc
Alexander Alekhin [Fri, 10 Jan 2020 11:38:17 +0000 (11:38 +0000)]
Merge pull request #16302 from sthibaul:dlopen-glibc

4 years agoMerge pull request #16300 from Palmitoxico:cmake-cmp0077
Alexander Alekhin [Wed, 8 Jan 2020 21:52:30 +0000 (21:52 +0000)]
Merge pull request #16300 from Palmitoxico:cmake-cmp0077

4 years agoEnable cmake CMP0077 policy
Augusto Fraga Giachero [Tue, 7 Jan 2020 11:36:35 +0000 (08:36 -0300)]
Enable cmake CMP0077 policy

When building OpenCV as a sub-project using cmake's add_subdirectory()
the OpenCV's build options would be overwritten to its default
state. With cmake 3.13+ the CMP0077 policy, option() honors previous
definitions via set().

4 years agoFix build on non-Linux glibc-based systems
Samuel Thibault [Wed, 8 Jan 2020 00:32:12 +0000 (01:32 +0100)]
Fix build on non-Linux glibc-based systems

dl functions are provided by all glibc-based systems (GNU/Linux, but
also GNU/Hurd, GNU/kFreeBSD)

4 years agoMerge pull request #16289 from Nuzhny007:3.4
Alexander Alekhin [Tue, 7 Jan 2020 22:13:39 +0000 (22:13 +0000)]
Merge pull request #16289 from Nuzhny007:3.4

4 years agoFixed compilation on windows with openvx
Nuzhny007 [Mon, 6 Jan 2020 03:32:56 +0000 (06:32 +0300)]
Fixed compilation on windows with openvx

4 years agoMerge pull request #16265 from JanuszL:fix_bgr
Alexander Alekhin [Sun, 5 Jan 2020 12:01:03 +0000 (12:01 +0000)]
Merge pull request #16265 from JanuszL:fix_bgr

4 years agoMerge pull request #16230 from YashasSamaga:cuda4dnn-fp-conversion
Alexander Alekhin [Sun, 5 Jan 2020 11:59:33 +0000 (11:59 +0000)]
Merge pull request #16230 from YashasSamaga:cuda4dnn-fp-conversion

4 years agoFix swapped channels in BGR* conversion utility function
Janusz Lisiecki [Fri, 20 Dec 2019 16:54:37 +0000 (17:54 +0100)]
Fix swapped channels in BGR* conversion utility function

- some of `icvCvt_BGR*` functions have R with B channels
  swapped what leads to the wrong conversion
- renames misleading `rgb` variable name to `bgr`
- swap back the conversion coefficients, `cB` should be the first

Signed-off-by: Janusz Lisiecki <jlisiecki@nvidia.com>
4 years agomake stitching script more pythonic
Ianaré Sévi [Sun, 29 Dec 2019 10:50:02 +0000 (11:50 +0100)]
make stitching script more pythonic

4 years agoimgproc(resize): improve 8u3 HResize vector exit calc
Paul E. Murphy [Thu, 12 Dec 2019 16:33:08 +0000 (10:33 -0600)]
imgproc(resize): improve 8u3 HResize vector exit calc

Actually, we can do this in constant time. xofs always
contains same or increasing offset values. We can instead
find the most extreme value used and never attempt to load it.

Similarly, we can note for all dx >= 0 and dx < (dwidth - cn)
where xofs[dx] + cn < xofs[dwidth-cn] implies dx < (dwidth - cn).

Thus, we can use this to control our loop termination optimally.

This fixes #16137 with little or no performance impact. I have
also added a debug check as a sanity check.

4 years agoMerge pull request #16238 from alalek:imgproc_resize_fix_types
Alexander Alekhin [Fri, 3 Jan 2020 16:30:28 +0000 (16:30 +0000)]
Merge pull request #16238 from alalek:imgproc_resize_fix_types

4 years agoMerge pull request #16222 from hannesa2:AndroidMoveVersionInfo
hannesa2 [Fri, 3 Jan 2020 16:15:35 +0000 (17:15 +0100)]
Merge pull request #16222 from hannesa2:AndroidMoveVersionInfo

Android move version info to Gradle style

* move version info to Gradle style

* Use VERSION_NAME in loader

* apply review suggestion

4 years agoMerge pull request #16275 from andybrucenet:fix_vector_insert
Andrew Bruce [Fri, 3 Jan 2020 16:09:40 +0000 (11:09 -0500)]
Merge pull request #16275 from andybrucenet:fix_vector_insert

* Ticket 16274: Fix compile error on std::vector.insert

* See https://github.com/opencv/opencv/issues/16274

4 years agocopyright: 2020
Alexander Alekhin [Thu, 2 Jan 2020 11:58:20 +0000 (11:58 +0000)]
copyright: 2020

4 years agocopyright: 2020
Alexander Alekhin [Thu, 2 Jan 2020 11:58:20 +0000 (11:58 +0000)]
copyright: 2020

4 years agoMerge remote-tracking branch 'upstream/3.4' into merge-3.4
Alexander Alekhin [Tue, 31 Dec 2019 10:11:39 +0000 (10:11 +0000)]
Merge remote-tracking branch 'upstream/3.4' into merge-3.4

4 years agoMerge pull request #16249 from alalek:check_size
Alexander Alekhin [Tue, 31 Dec 2019 10:09:00 +0000 (10:09 +0000)]
Merge pull request #16249 from alalek:check_size

4 years agoSort text TensorFlow graphs
Dmitry Kurtaev [Tue, 31 Dec 2019 08:43:32 +0000 (11:43 +0300)]
Sort text TensorFlow graphs

4 years agoMerge pull request #16264 from YashasSamaga:cuda4dnn-msvc-v140-workaround
Alexander Alekhin [Tue, 31 Dec 2019 08:20:24 +0000 (08:20 +0000)]
Merge pull request #16264 from YashasSamaga:cuda4dnn-msvc-v140-workaround

4 years agoMerge pull request #16263 from dkurt:dnn_tf_addv2
Alexander Alekhin [Mon, 30 Dec 2019 22:56:11 +0000 (22:56 +0000)]
Merge pull request #16263 from dkurt:dnn_tf_addv2

4 years agoMerge pull request #16258 from jed-frey:patch/python38-clock
Alexander Alekhin [Mon, 30 Dec 2019 17:34:23 +0000 (17:34 +0000)]
Merge pull request #16258 from jed-frey:patch/python38-clock

4 years agosimplify code to help MSVC 19.10 and lower
YashasSamaga [Mon, 30 Dec 2019 17:32:17 +0000 (23:02 +0530)]
simplify code to help MSVC 19.10 and lower

4 years agoAddV2 from TensorFlow
Dmitry Kurtaev [Mon, 30 Dec 2019 17:06:58 +0000 (20:06 +0300)]
AddV2 from TensorFlow