opencv

Commit Graph

Author	SHA1	Message	Date
Sayed Adel	9ea62bfddb	core:vsx reimplement `v_broadcast_element()` There's no need to use `vec_perm()` instead of `vec_splat()`, since instruction `vperm` is quite heavy compared to `vsplt[b,h,w]`.	5 years ago
Alexander Alekhin	4e56c1326f	core: adjust type of allocator_stats counter, allow to disable	5 years ago
Alexander Alekhin	3c85745e22	pre: OpenCV 4.3.0 (version++)	5 years ago
Alexander Alekhin	b7ecaceda8	pre: OpenCV 3.4.10 (version++) - Android Manager version is not increased (stuck on 3.49)	5 years ago
Alexander Alekhin	3a2f40ac6f	core: don't allow reallocation in add/div/sub/bitwise aug operators	5 years ago
Manoj Gupta	880d2afb67	Fix building with ToT libc++ ToT libc++ (LLVM) no longer includes <sstream> as part of <complex> which breaks building opencv. Include <sstream> header explcitly to fix this.	5 years ago
Alexander Alekhin	4f288a1e28	Merge pull request #16704 from alalek:core_log_once_log_if * core(logger): add CV_LOG_ONCE_xxx() CV_LOG_IF_xxx() macros * core(logger): keep tests disabled	5 years ago
Peter Würtz	5012fc5d23	Merge pull request #16684 from pwuertz:ignore_clang_mat_inl * Ignore clang warnings for deprecated enum+enum operations in mat.inl.hpp * build: added customization macros, cmake flags for OpenCV build	5 years ago
Alexander Alekhin	84bb67e49b	exclude opencv_contrib modules	5 years ago
Peter Würtz	53c0189ed7	Allow access to CUDA memory and stream pointers for interoperability.	5 years ago
Maksim Shabunin	bf96d8239d	Use BufferArea in more places	5 years ago
Alexander Alekhin	01048e5603	Merge pull request #16616 from alalek:dnn_fix_input_shape * dnn: fix processing of input shapes - importer: avoid using of .setInput() => .setInputShape() - setInput: shape limitation check (partial) * dnn(test): test .setInput() in readNet()	5 years ago
Maksim Shabunin	8b2c499be6	intrin: fixed int64->double conversion for AVX-512	5 years ago
Alexander Smorkalov	c87b99e82b	Added test for new MatX division.	5 years ago
Pavel Rojtberg	e13a73d084	core: export getCPUFeaturesLine to bindings	5 years ago
Maksim Shabunin	55cdeaa6dd	BufferArea: initial version, usage in StereoBM New class BufferArea is used to hide complexity of buffers allocations and allow instrumentation with valgrind and sanitizers.	5 years ago
gapry	ac9f8c1f41	Fixed Compilation warnings \| Issue #16336	5 years ago
Alexander Alekhin	a4bd7506a5	core: CV_STRONG_ALIGNMENT macro Should be used to guard unsafe type casts of pointers	5 years ago
Niklas Hambüchen	70cbc3d883	cvdef.h: Don't use C's limits.h under C++ Just like with the other headers in the rest of the file. See e.g. https://stackoverflow.com/questions/36831465/what-difference-does-it-make-when-i-include-limits-or-limits-h-in-my-c-cod for the reasons, the most important one being that limits.h does not respect namespaces, which can make problems for downstream consumers of cvdef.h.	5 years ago
Sayed Adel	ec033330df	core:vsx workaround for the unexpected results of `vec_vbpermq` in gcc4.9	5 years ago
Sayed Adel	bd531bd828	core:vsx fix inline asm constraints generalize constraints to 'wa' for VSX registers	5 years ago
Alexander Alekhin	d42e04d0df	core(SIMD): fix MSA build - add v_reduce_min/max for u8/s8	5 years ago
Chip Kerchner	301626ba26	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	5 years ago
Alexander Alekhin	a9f3acb125	core(simd): fix NEON alignmnet issue	5 years ago
Alexander Alekhin	e180cc050b	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	5 years ago
Alexander Alekhin	523f081923	core(check): add Size_<int>	5 years ago
Brian Wignall	f9c514b391	Fix spelling typos backport commit `659ffaddb4`	5 years ago
Brian Wignall	659ffaddb4	Fix spelling typos	5 years ago
Alexander Alekhin	bda89a6469	release: OpenCV 4.2.0	5 years ago
Alexander Alekhin	64e6cf9fe5	release: OpenCV 3.4.9	5 years ago
Alexander Alekhin	8d22ac200f	core: workaround flipHoriz() alignment issues	5 years ago
Tatsuro Shibamura	971ae00942	Merge pull request #16027 from shibayan:arm64-windows10 * Support ARM64 Windows 10 platform * Fixed detection issue for ARM64 Windows 10 * Try enabling ARM NEON intrin * build: disable NEON with MSVC compiler * samples(directx): gdi32 dependency	5 years ago
Alexander Alekhin	416848066c	core: provide safe implementations of CV_XADD() only	5 years ago
Alexander Alekhin	76b5e19eb3	core: add "namespace cv" in CV_StaticAssert fallback implementation	5 years ago
Alexander Alekhin	a675c4937a	core: OPENCV_INCLUDE_PORT_FILE for custom platform configuration	5 years ago
Paul Murphy	a011035ed6	Merge pull request #15257 from pmur:resize * resize: HResizeLinear reduce duplicate work There appears to be a 2x unroll of the HResizeLinear against k, however the k value is only incremented by 1 during the unroll. This results in k - 1 duplicate passes when k > 1. Likewise, the final pass may not respect the work done by the vector loop. Start it with the offset returned by the vector op if implemented. Note, no vector ops are implemented today. The performance is most noticable on a linear downscale. A set of performance tests are added to characterize this. The performance improvement is 10-50% depending on the scaling. * imgproc: vectorize HResizeLinear Performance is mostly gated by the gather operations for x inputs. Likewise, provide a 2x unroll against k, this reduces the number of alpha gathers by 1/2 for larger k. While not a 4x improvement, it still performs substantially better under P9 for a 1.4x improvement. P8 baseline is 1.05-1.10x due to reduced VSX instruction set. For float types, this results in a more modest 1.2x improvement. * Update U8 processing for non-bitexact linear resize * core: hal: vsx: improve v_load_expand_q With a little help, we can do this quickly without gprs on all VSX enabled targets. * resize: Fix cn == 3 step per feedback Per feedback, ensure we don't overrun. This was caught via the failure observed in Test_TensorFlow.inception_accuracy.	5 years ago
Alexander Alekhin	c3023fb52b	pre: OpenCV 4.2.0 (version++)	5 years ago
Alexander Alekhin	76a27e3399	pre: OpenCV 3.4.9 (version++)	5 years ago
Alexander Alekhin	f21bde4d9f	Merge pull request #16046 from alalek:issue_15990 * core: disable invalid constructors in C API by default - C API objects will lose their default initializers through constructors * samples: stop using of C API	5 years ago
Vadim Levin	8d74101f07	Merge pull request #15955 from VadimLevin:dev/vlevin/generator_tests Tests for argument conversion of Python bindings generator * Tests for parsing elemental types from Python bindings - Add positive and negative tests for int, float, double, size_t, const char, bool. - Tests with wrong conversion behavior are skipped. Move implicit conversion of bool to integer/floating types to wrong conversion behavior.	5 years ago
Brian Wignall	af997529a1	Fix some typos	5 years ago
Brian Wignall	9276f1910b	Fix some typos	5 years ago
Alexander Alekhin	ec55b6f6db	core: fix MSA build	5 years ago
Everton Constantino	75315fb297	Merge pull request #15494 from everton1984:hal_vector_get_n Improving VSX performance of integral function * Adding support for vector get function on VSX datatypes so the integral function gains a bit of performance. * Removing get as a datatype member function and implementing a new HAL instruction v_extract_n to get the n-th element of a vector register. * Adding SSE/NEON/AVX intrinsics. * Implement new HAL instruction v_broadcast_element on VSX/AVX/NEON/SSE. * core(simd): add tests for v_extract_n/v_broadcast_element - updated docs - commented out code to repair compilation - added WASM and MSA default implementations * core(simd): fix compilation - x86: avoid _mm256_extract_epi64/32/16/8 with MSVS 2015 - x86: _mm_extract_epi64 is 64-bit only * cleanup	5 years ago
Alexander Alekhin	e07a488012	Merge pull request #15925 from alalek:core_test_simd_cpp_emulation core(test): extending tests with SIMD C++ emulation code (intrin_cpp.hpp) * core(test): test SIMD CPP emulation code (intrin_cpp.hpp) * core(simd): eliminate build warnings from intrin_cpp.hpp	5 years ago
Christoph Bachhuber	c638f085aa	Refactor for clarity and avoiding code duplication Implement GArik's comments Remove unnecessary c_str() Fix brace position	5 years ago
Alexander Alekhin	7ecdcf6ca6	build: GCC9 compilation	5 years ago
Igor Murzov	6d5b900324	Simplify OpenCL info dumping code: * Reduce code nesting * Drop redundant .c_str() calls	5 years ago
Alexander Alekhin	a893969ec9	core(simd): v_setall template	5 years ago
Igor Murzov	a9d23a6479	Fix wording in some tutorials	5 years ago

... 2 3 4 5 6 ...

2266 Commits (be110d04640cffdc8ac9713d7db02da2e3033400)