Commit Graph

3322 Commits

Author SHA1 Message Date
Alexander Smorkalov
1a6f669763 Merge branch 4.x 2026-04-09 18:44:48 +03:00
Jorge Velez
a49a293d3c Merge pull request #27534 from JorgeV92:gsoc2025-tokenizer
GSoC 2025: Add Tokenizer Support to DNN Module #27534

merge with https://github.com/opencv/opencv_extra/pull/1276

### Summary
This pull request introduces initial support for a tokenizer module under `modules/dnn/src/tokenizer` as part of Google Summer of Code 2025 (Project: Tokenization for OpenCV DNN).

### Status
- [x] Project structure in place
- [x] Initial BPE tokenizer loading
- [x] Regex splitting (in progress)
- [x] Encoding logic for GPT-2 tokenizer (in progress)
- [ ] Documentation (to be improved)

### Goals
The goal is to support Hugging Face-compatible tokenization (e.g., GPT-2) natively in C++ to be integrated with DNN inference pipelines. 

The core pipeline lives in `dnn/src/tokenizer/core_bpe.hpp` and `dnn/src/tokenizer/encoding.hpp`. For Unicode handling I’m using `dnn/src/tokenizer/unicode.hpp`, which is adapted from llama.cpp.


### Feedback
Please share early feedback on:
- General design structure
- Integration strategy with `dnn`
- Code organization or naming conventions

### Reference
Project: https://summerofcode.withgoogle.com/programs/2025/projects/79SW6eNK
2026-04-06 10:46:13 +03:00
Lurie97
a3e129aad8 Merge pull request #28686 from Lurie97:fix_inplace
core(opencl): fix inplace transpose race by enforcing LLSS ordering via local barrier #28686

The former inplace transpose implementation allowed a reordering of global-memory operations across work-items. Specifically, the intended LLSS (Load–Load–Store–Store) access pattern could be reordered by the GPU into LSLS (Load–Store–Load–Store), causing partially written tiles to be observed by other work-items and producing incorrect output.

This patch introduces a tiled LDS-based algorithm and adds an explicit:

    barrier(CLK_LOCAL_MEM_FENCE);

between the load and store phases.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [ ] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-04-03 15:50:11 +03:00
Prasad Ayush Kumar
47ac80995f Merge pull request #28749 from Prasadayus:NMS-empty-detection-fix
Nms empty detection fix #28749

Requires opencv_extra: https://github.com/opencv/opencv_extra/pull/1330

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-04-02 18:37:23 +03:00
pratham-mcw
3cf98c51c8 Merge pull request #28609 from pratham-mcw:core-rotate-neon-optimization
core: add NEON implementation for rotate function #28609

- This PR adds a NEON intrinsics-based implementation for the rotate function in matrix_transform.cpp for Windows-ARM64.
- The optimized implementation uses  ARM NEON intrinsics to accelerate the internal transpose step used by the rotate function.
- In the x64 architecture, the rotate operation benefits from IPP-based optimized implementations. However, on ARM64, the execution falls back to the scalar implementation, which results in lower performance.
- To achieve performance parity with x64, a NEON-based SIMD implementation has been added for ARM64. 
- After introducing these changes, the rotate function showed noticeable performance improvements on ARM64 platforms.
<img width="1009" height="817" alt="image" src="https://github.com/user-attachments/assets/8bec0041-b19c-4fc8-9103-532746224515" />

 
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
2026-03-31 16:52:04 +03:00
Alexander Smorkalov
92c43f80fc Merge pull request #28739 from pratham-mcw:core/fix-MeanStdDeviation-accuracytest
core: fix meanStdDev bug by using separate variables v2, v3 in sumsqr_
2026-03-31 15:39:16 +03:00
pratham-mcw
c5d747f75c core: fix meanStdDev bug by using separate variables v2, v3 in sumsqr_ 2026-03-31 11:29:34 +05:30
Vincent Rabaud
5c91261ca0 Force step to be ptrdiff_t in resize
Otherwise, ASAN could return an error:
"runtime error: addition of unsigned offset"
2026-03-30 10:48:00 +02:00
Yang Guanyuhan
2d15169ab5 Merge pull request #28692 from YangGuanyuhan:fix-lightglue-assertion-fail
dnn: fix dst_dp assertion in broadcast for size-1 dims causing crash in lightglue.onnx model #28692

The original assertion CV_Assert(dst_dp == 1) does not handle valid cases where the innermost dimension size is 1 like [10, 5, 1], resulting in dst_dp == 0.

This occurs during broadcasting in LightGlue ONNX model and leads to assertion failure.

Allow dst_dp == 0 for size-1 dimensions to handle this edge case correctly.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-03-23 11:27:37 +03:00
Abhishek Gola
ff2e6358fd added AVXX VNNI support 2026-03-20 13:20:31 +05:30
Alexander Smorkalov
f6aceee13b Merge pull request #28655 from usernotfound-101:cvmixchannels-warning-fix
Add static integer casting for cvMixChannels, safer
2026-03-20 10:25:38 +03:00
Varun Jaiswal
df7d437b86 Merge pull request #28663 from varun-jaiswal17:added-inRange-64double-SIMD
Added SIMD support for 64 bit float and fallback for 64 bit int#28663

Merged with : https://github.com/opencv/ci-gha-workflow/pull/299

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-03-20 08:42:56 +03:00
Alexander Smorkalov
b5a7e0c662 Merge branch 4.x 2026-03-19 11:29:57 +03:00
Abhishek Gola
2dec80044b moved AVXVNNI function to core 2026-03-17 18:58:03 +05:30
usernotfound-101
3470f5b35b Add static integer casting to get rid of the warning, safer 2026-03-13 22:25:08 +05:30
Vadim Pisarevsky
1b483ffea6 Merge pull request #28585 from vpisarev:dnn_block_layout_v5
Block layout-based convolution in DNN #28585

merge together with https://github.com/opencv/opencv_extra/pull/1321

Some core parts of the new engine in DNN module have been revised substantially:

1. all tests seem to pass, except for `Test_Graph_Simplifier.ResizeSubgraph`, which has been disabled because it does not take the newly added `TransformLayoutLayer` into account. The test should be reworked perhaps.
1. convolution and related operations (maxpool/avgpool) now use so-called block layout (`DATA_LAYOUT_BLOCK`), where `NxCxHxW` tensors are represented  as `NxC1xHxWxC0`, where `C1=(C + C0-1)/C0` and `C0` is a power-of-two (usually 4, 8, 16 or 32).
1. graph is now pre-processed and `TransformLayoutLayer` is inserted to convert data from NCHW or NHWC layout to the block layout or vice versa. The transformations are done in a lazy way only when they are really needed. For example, in the whole Resnet only 2 transformations are performed.
1. transformer-based models and other models that do not use convolutions will run as usual, without going to block layout.
1. there is yet another graph preprocessing stage added that embeds constant weights/scale and bias into convolution and batch norm layers.
1. 'batchnorm', 'activation' and 'adding a residual' are now fused with convolution, just like in the old engine. That brings some noticeable acceleration.
1. optimized convolution kernels have been added.
     * depthwise convolution, as well as maxpool and avgpool support C0=4, 8, 16 etc. _as long as_  C0 is divisible by the number of fp32 lanes in a SIMD register of the target platform (e.g. on ARM with NEON there must be `C0 % 4 == 0`, on x64 with AVX2 `C0 % 8 == 0`).
     * non-depthwise convolution only supports C0=8 for now. C0=8 seems to be a sweetspot for ARM with NEON, x64 with AVX2 or RISC-V with RVV (with 128- or 256-bit registers). For some platforms with dedicated matrix accelerators C0=16 or even C0=32 might be more efficient, but we could add the respective kernels later.
     * only fp32 kernels have been added. fp16/bf16 kernels might be added a little later.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
- [x] Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2026-03-13 17:09:27 +03:00
Madan mohan Manokar
18c7c9bcb9 Merge pull request #28614 from amd:fast_flipHoriz
Optimized flip Horizontal #28614

- Refactor flipHoriz implementation.
- Optimizations for horizontal image flipping added.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-03-13 11:11:20 +03:00
Matt Van Horn
8e1fa1bbdc Merge pull request #28620 from mvanhorn:osc/28619-fix-yaml-parsekey-empty-key-oob
core: fix heap-buffer-overflow in YAML parseKey for empty keys #28620

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake

### Description

Fixes https://github.com/opencv/opencv/issues/28619

Moves the "empty key" check before the backward do-while scan in `YAMLParser::parseKey()`.

**Problem:** When parsing a YAML mapping with an empty key (e.g. `: 10` at column 0), `endptr == ptr` after the forward scan finds `:`. The do-while loop `do c = *--endptr; while(c == ' ')` always executes at least once, so it decrements `endptr` to `ptr-1` and reads one byte before the heap allocation (ASan: heap-buffer-overflow READ of size 1).

**Fix:** Check `endptr == ptr` before entering the backward loop. If the key is empty, raise `CV_PARSE_ERROR_CPP("An empty key")` immediately without the OOB read.

This contribution was developed with AI assistance (Claude Code).
2026-03-12 10:37:24 +03:00
Alexander Smorkalov
1a3e1e05b7 Merge pull request #28581 from JoyBoy900908:fix-ubsan-countnonzero
core: fix UBSan function pointer type mismatch in countNonZero
2026-03-12 10:16:17 +03:00
JoyBoy900908
7286212ce6 core: use void* in countNonZero signature per team review 2026-03-09 14:20:43 +08:00
Alexander Smorkalov
f46498b3c7 Merge pull request #28587 from intel-staging:staging/ekharkov/arithm
Restored IPP in cv::compare
2026-03-03 08:49:20 +03:00
ekharkov
30aa3c9266 Revert "Disable IPP with AVX512 in cv::compare because of performance regression" 2026-03-02 01:47:14 -08:00
Abhishek Gola
31cab2a055 Merge pull request #28574 from abhishek-gola:reduce_layer_empty_set
* added empty set support

* accept unnamed dynamic dims

* Fix for LSTM test failure

* removed converToND

* openvino failing test fix

* ARM CI issue fix

* ARM issue fix
2026-03-01 15:19:44 +03:00
JoyBoy900908
a9e13a5e8c core: fix UBSan function pointer type mismatch in countNonZero 2026-02-28 20:50:54 +08:00
Alexander Smorkalov
01320ab612 Fixed cv::mul overflow for U16 type. 2026-02-27 10:04:59 +03:00
Alexander Smorkalov
6c995c768f Merge branch 4.x 2026-02-20 22:05:35 +03:00
Sergey Fedorov
e1e170afc2 parallel_impl.cpp: fix assembler syntax for powerpc*-darwin 2026-02-18 04:35:41 +08:00
Adrian Kretz
196a8afe76 Merge pull request #28267 from akretz:absdiff-int
Fix absdiff with int arguments #28267

I believe the fix to the undefined behavior described in #27080 is simply casting to unsigned before subtraction, because

1. casting int to unsigned is well-defined; negative values get represented modulo $2^{32}$
2. overflow in unsigned subtraction is well-defined and the results are modulo $2^{32}$

Since we are computing everything modulo $2^{32}$ and the result must always be a non-negative number below $2^{32}$, this computation should be well-defined and correct.

I have verified this on ARM Apple Clang and on x64 Linux gcc with `-O3` and both produce the correct values in the reproducer of #27080. The test I have added fails with the integer overflow on both platforms I have tested. Perhaps @fengyuentau could verify if this also fixes the issue on the platforms he has tested?

I have added a fix and removed the workarounds. I recommended this approach in #28229, but he decided to revert it.


### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2026-02-15 10:29:29 +03:00
Alexander Smorkalov
82ff8e45e9 Merge branch 4.x 2026-02-14 15:37:33 +03:00
Erellu
b201018a25 Fix UB in cv::error when breakOnError set to true
Fixes #28436.
Replaces a nullptr dereference (unguaranteed to cause a SEGFAULT) by `std::terminate` which will call the current `std::terminate_handler` and then SIGARBT.
2026-01-22 13:26:02 +01:00
Vincent Rabaud
ef0172d636 Improve precision of RotatedRect::points
Not by much but enough to get a test to pass.
2026-01-08 09:52:35 +01:00
WalkingDevFlag
175dd57bdd Fix copyTo on empty fixed-type matrices (#28343)
PR #27972 added _dst.create(size(), type()) in copyTo's empty() block.
In Debug builds, Mat::release() was resetting flags to MAGIC_VAL,
clearing the type information and causing assertion failures when
destination has fixedType().

Preserve type flags in Mat::release() debug mode by using:
  flags = (flags & CV_MAT_TYPE_MASK) | MAGIC_VAL

Thanks to @akretz for suggesting this better approach.
2026-01-06 00:51:06 +05:30
Vincent Rabaud
5622958189 Replace pow with std::pow
The C pow casts to double while std::pow has overloads that can be
optimized by the compiler.
Also replace pow(*, 1./3) by cbrt.
2026-01-04 15:12:59 +01:00
Alexander Smorkalov
2b1cf03a9d Disable IPP with AVX512 in cv::compare because of performance regression. 2025-12-26 09:19:28 +03:00
Vadim Pisarevsky
eaccbe24b2 Merge pull request #28242 from vpisarev:fix_input_array_std_vector
modified Input/OutputArray methods to handle 'std::vector<T>' or 'std::vector<std::vector<T>>' properly #28242

This is port of #26408 with some further improvements (all switch-by-vector-type statements are consolidated in a single macro)

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [ ] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2025-12-23 14:25:07 +03:00
satyam yadav
74fbfc2343 Merge pull request #28117 from satyam102006:fix/solveCubic-numerical-instability
core: fix solveCubic numerical instability via coefficient normalization (fixes #27748) #28117

 Summary
This PR fixes numerical instability in `cv::solveCubic` when the leading coefficient `a` is non-zero but extremely small relative to other coefficients (Issue #27748). 

It introduces a **normalization step** that scales all coefficients by their maximum magnitude before solving. This ensures robust detection of when the equation should degenerate to a quadratic solver, without breaking valid cubic equations that happen to have small coefficients (e.g., scaled by 1e-9).

The Problem (Issue #27748)
The previous implementation checked `if (a == 0)` to decide whether to use the cubic or quadratic formula. 
- When `a` is extremely small (e.g., 1e-17) but not exactly zero, and other coefficients are normal (e.g., 5.0), the standard cubic formula suffers from catastrophic cancellation and overflow, producing incorrect roots (e.g., 1e14).

 The Fix
1. Normalization: The solver now finds `max_coeff = max(|a|, |b|, |c|, |d|)` and scales all coefficients by `1.0 / max_coeff`.
2. Relative Threshold: It then checks `if (abs(a) < epsilon)` on the *normalized* coefficients. 

 Why this is better than previous attempts
In a previous attempt (PR #28057), a simple absolute check `abs(a) < epsilon` was proposed. That approach was rejected because it failed for scaled equations. 


Fixes #27748
2025-12-22 11:01:51 +03:00
Brian Ferri
93f1413db9 fix: Invalid operands to binary expression 2025-12-19 10:49:36 +01:00
AdwaithBatchu
87f4a9782c Merge pull request #28203 from AdwaithBatchu:fix-macos-lapack-warnings
core: suppress LAPACK deprecation warnings on macOS #28203

### Description
On macOS (Apple Silicon) with newer Xcode/Clang, the CLAPACK interface is deprecated.
Compiling `hal_internal.cpp` triggers multiple warnings like:
`warning: 'sgesv_' is deprecated: first deprecated in macOS 13.3 - The CLAPACK interface is deprecated. [-Wdeprecated-declarations]`

Since migrating to the new Accelerate interface (Apple's recommended fix) requires significant changes to the HAL implementation and breaks the build if headers are mismatched, this PR takes the pragmatic approach. It suppresses the `-Wdeprecated-declarations` warning for this specific file using Clang pragmas to clean up the build output.

### Verification
- [x] Verified build is silent on macOS (M3).
- [x] Verified pragmas are correctly scoped within `HAVE_LAPACK` to prevent scope mismatch on non-LAPACK builds.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2025-12-18 13:53:56 +03:00
Abhishek Gola
111354bfef Added support for vector of gpuMatND 2025-12-10 16:12:48 +05:30
Vincent Rabaud
01f2f3c4b9 Disable MSAn for lapack_LU
One of the arguments is sent to Fortran so the whole function has
to be disabled fro MSAN. Arguments cannot be unpoisoned, cf
https://g3doc.corp.google.com/testing/msan/g3doc/index.md?cl=head#memorysanitizer-api
2025-12-08 21:53:53 +01:00
Vincent Rabaud
c6220c2b47 Add missing combaseapi.h include.
This is needed for CoCreateGuid.
2025-12-08 08:53:54 +01:00
Alexander Smorkalov
fd11f635c1 Merge pull request #28122 from asmorkalov:as/dynamic_openmp
Dropped OPENCV_FOR_OPENMP_DYNAMIC_DISABLE environment varibale in favor of standard OMP_DYNAMIC #28122

Fixes: https://github.com/opencv/opencv/issues/25717
Replaces: https://github.com/opencv/opencv/pull/28084

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [ ] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2025-12-05 14:13:19 +03:00
nishith-fujitsu
8efc0fd47b Merge pull request #28055 from nishith-fujitsu:sve_fastGEMM1t
dnn: add SVE optimized fastGEMM1T function and SVE dispatch #28055

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch

**Description**
This PR enables fastGemm1t vectorized with SVE for AARCH64 architecture that called by recurrent layers and fully connected layers with SVE dispatching mechanism.

**ARM Compatibility:**
Modified the build scripts, and configuration files to ensure compatibility with ARM processors.

**Checklist**

Code changes have been tested on ARM devices (Graviton3).

**Modifications**

- Implemented FastGemm1T kernel in SVE with Vector length agnostic approach.

- Added Flags and checks to call our ported Kernel in Recurrent Layer and FullyConnected layer.

- Changes made to cmakelist.txt to dispatch our ported kernel for SVE.

- Flag OpenCV Dispatch with SVE optimization is added to support SVE implemented kernel for OpenCV. According to OpenCV build optimization https://github.com/opencv/opencv/wiki/CPU-optimizations-build-options 
cmake \
    -DCPU_BASELINE=NEON\
    -D CPU_DISPATCH=SVE\

**Performance Improvement**
- The suggested optimizations Improves the performance of LSTM layer and fully connected layer.
<html xmlns:o="urn:schemas-microsoft-com:office:office"
xmlns:x="urn:schemas-microsoft-com:office:excel"
xmlns="http://www.w3.org/TR/REC-html40">

<head>

<meta name=ProgId content=Excel.Sheet>
<meta name=Generator content="Microsoft Excel 15">
<link id=Main-File rel=Main-File
href="file:///C:/Users/jaiswaln/AppData/Local/Temp/msohtmlclip1/01/clip.htm">
<link rel=File-List
href="file:///C:/Users/jaiswaln/AppData/Local/Temp/msohtmlclip1/01/clip_filelist.xml">
<style>
<!--table
	{mso-displayed-decimal-separator:"\.";
	mso-displayed-thousand-separator:"\,";}
@page
	{margin:.75in .7in .75in .7in;
	mso-header-margin:.3in;
	mso-footer-margin:.3in;}
tr
	{mso-height-source:auto;}
col
	{mso-width-source:auto;}
br
	{mso-data-placement:same-cell;}
td
	{padding-top:1px;
	padding-right:1px;
	padding-left:1px;
	mso-ignore:padding;
	color:black;
	font-size:11.0pt;
	font-weight:400;
	font-style:normal;
	text-decoration:none;
	font-family:"Aptos Narrow", sans-serif;
	mso-font-charset:0;
	mso-number-format:General;
	text-align:general;
	vertical-align:bottom;
	border:none;
	mso-background-source:auto;
	mso-pattern:auto;
	mso-protection:locked visible;
	white-space:nowrap;
	mso-rotate:0;}
.xl63
	{border:.5pt solid windowtext;}
.xl64
	{text-align:center;}
.xl65
	{text-align:center;
	border:.5pt solid windowtext;}
-->
</style>
</head>

<body link="#467886" vlink="#96607D">


Name of Test | dnn_neon | dnn_sve | dnn_sve   vs dnn_neon(x-factor)
-- | -- | -- | --
lstm::Layer_LSTM::BATCH=1,   IN=64, HIDDEN=192, TS=100 | 2.878 | 2.326 | 1.24
lstm::Layer_LSTM::BATCH=1,   IN=192, HIDDEN=192, TS=100 | 4.162 | 3.08 | 1.35
lstm::Layer_LSTM::BATCH=1,   IN=192, HIDDEN=512, TS=100 | 18.627 | 16.152 | 1.15
lstm::Layer_LSTM::BATCH=1,   IN=1024, HIDDEN=192, TS=100 | 10.98 | 7.976 | 1.38
lstm::Layer_LSTM::BATCH=64,   IN=64, HIDDEN=192, TS=2 | 4.41 | 3.459 | 1.27
lstm::Layer_LSTM::BATCH=64,   IN=192, HIDDEN=192, TS=2 | 6.567 | 4.807 | 1.37
lstm::Layer_LSTM::BATCH=64,   IN=192, HIDDEN=512, TS=2 | 28.471 | 22.909 | 1.24
lstm::Layer_LSTM::BATCH=64,   IN=1024, HIDDEN=192, TS=2 | 15.491 | 12.537 | 1.24
lstm::Layer_LSTM::BATCH=128,   IN=64, HIDDEN=192, TS=2 | 8.848 | 6.821 | 1.3
lstm::Layer_LSTM::BATCH=128,   IN=192, HIDDEN=192, TS=2 | 12.969 | 9.522 | 1.36
lstm::Layer_LSTM::BATCH=128,   IN=192, HIDDEN=512, TS=2 | 55.52 | 45.746 | 1.21
lstm::Layer_LSTM::BATCH=128,   IN=1024, HIDDEN=192, TS=2 | 31.226 | 26.132 | 1.19

</body>

</html>

<html xmlns:v="urn:schemas-microsoft-com:vml"
xmlns:o="urn:schemas-microsoft-com:office:office"
xmlns:x="urn:schemas-microsoft-com:office:excel"
xmlns="http://www.w3.org/TR/REC-html40">

<head>

<meta name=ProgId content=Excel.Sheet>
<meta name=Generator content="Microsoft Excel 15">
<link id=Main-File rel=Main-File
href="file:///C:/Users/jaiswaln/AppData/Local/Temp/msohtmlclip1/01/clip.htm">
<link rel=File-List
href="file:///C:/Users/jaiswaln/AppData/Local/Temp/msohtmlclip1/01/clip_filelist.xml">
<style>
<!--table
	{mso-displayed-decimal-separator:"\.";
	mso-displayed-thousand-separator:"\,";}
@page
	{margin:.75in .7in .75in .7in;
	mso-header-margin:.3in;
	mso-footer-margin:.3in;}
tr
	{mso-height-source:auto;}
col
	{mso-width-source:auto;}
br
	{mso-data-placement:same-cell;}
td
	{padding-top:1px;
	padding-right:1px;
	padding-left:1px;
	mso-ignore:padding;
	color:black;
	font-size:11.0pt;
	font-weight:400;
	font-style:normal;
	text-decoration:none;
	font-family:"Aptos Narrow", sans-serif;
	mso-font-charset:0;
	mso-number-format:General;
	text-align:general;
	vertical-align:bottom;
	border:none;
	mso-background-source:auto;
	mso-pattern:auto;
	mso-protection:locked visible;
	white-space:nowrap;
	mso-rotate:0;}
.xl65
	{border:.5pt solid windowtext;}
.xl66
	{text-align:center;}
.xl67
	{text-align:center;
	border:.5pt solid windowtext;}
-->
</style>
</head>

<body link="#467886" vlink="#96607D">


Name of Test | dnn_neon | dnn_sve | dnn_sve   vs dnn_neon(x-factor)
-- | -- | -- | --
fc::Layer_FullyConnected::([5,   16, 512, 128], 256, false, OCV/CPU) | 5.086 | 4.483 | 1.13
fc::Layer_FullyConnected::([5,   16, 512, 128], 256, true, OCV/CPU) | 8.512 | 8.347 | 1.02
fc::Layer_FullyConnected::([5,   16, 512, 128], 512, false, OCV/CPU) | 9.467 | 8.965 | 1.06
fc::Layer_FullyConnected::([5,   16, 512, 128], 512, true, OCV/CPU) | 14.855 | 13.527 | 1.1
fc::Layer_FullyConnected::([5,   16, 512, 128], 1024, false, OCV/CPU) | 18.821 | 18.023 | 1.04
fc::Layer_FullyConnected::([5,   16, 512, 128], 1024, true, OCV/CPU) | 27.558 | 24.966 | 1.1
fc::Layer_FullyConnected::([5,   512, 384, 0], 256, false, OCV/CPU) | 0.924 | 0.804 | 1.15
fc::Layer_FullyConnected::([5,   512, 384, 0], 256, true, OCV/CPU) | 1.259 | 1.126 | 1.12
fc::Layer_FullyConnected::([5,   512, 384, 0], 512, false, OCV/CPU) | 1.957 | 1.655 | 1.18
fc::Layer_FullyConnected::([5,   512, 384, 0], 512, true, OCV/CPU) | 2.831 | 2.775 | 1.02
fc::Layer_FullyConnected::([5,   512, 384, 0], 1024, false, OCV/CPU) | 5.92 | 6.379 | 0.93
fc::Layer_FullyConnected::([5,   512, 384, 0], 1024, true, OCV/CPU) | 8.924 | 8.993 | 0.99

</body>

</html>
2025-12-03 10:42:28 +03:00
Alexander Smorkalov
e466110245 Merge branch 4.x 2025-12-02 15:38:19 +03:00
Ghazi-raad
2cf9c68da0 Merge pull request #28087 from Ghazi-raad:fix/tempfile-race-condition-19648
Fix tempfile race condition on Windows (issue #19648) #28087

Fix tempfile race condition on Windows

Addresses issue #19648

Problem
The cv::tempfile() function on Windows used GetTempFileNameA() followed by an immediate DeleteFileA() call. This created a race condition where multiple OpenCV processes running simultaneously could receive the same temporary filename, leading to name collisions.

Root Cause
The previous implementation:

Called GetTempFileNameA() to generate a temp filename
Immediately deleted the file to free the name
Returned just the filename string
Between steps 2 and 3, another process could call GetTempFileNameA() and receive the same filename, causing a collision.

Solution
Replaced GetTempFileNameA() with GUID-based filename generation using CoCreateGuid(), following the same approach already used in GetTempFileNameWinRT() and Microsoft's recommendations for scenarios requiring many temp files.

Changes
modules/core/src/system.cpp:

Removed GetTempFileNameA() and DeleteFileA() calls
Added CoCreateGuid() to generate unique GUID-based filenames
Format: "ocv{GUID}" where GUID ensures uniqueness across processes
Benefits

Eliminates race condition in multi-process scenarios
No file I/O overhead from creating and deleting placeholder files
Consistent with WinRT implementation approach
Follows Microsoft best practices
Testing
Standard OpenCV test suite. The change only affects Windows temp file naming and maintains the same String return type and usage pattern.
2025-11-27 17:57:42 +03:00
Pierre Chatelier
fdf0332954 Merge pull request #23913 from chacha21:GpuMatND_InputOutputArray
make cuda::GpuMatND compatible with InputArray/OutputArray #23913

continuation of  [PR#19259](https://github.com/opencv/opencv/pull/19259) 

Make cuda::GpuMatND wrappable in InputArray/OutputArray
The goal for now is just wrapping, some functions are not supported (InputArray::size(), InputArray::convertTo(), InputArray::assign()...)

No new feature for cuda::GpuMatND

- [X] I agree to contribute to the project under Apache 2 License.
- [X] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [X] The PR is proposed to the proper branch
- [X] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2025-11-21 08:41:12 +03:00
Alexander Smorkalov
1b6fb61b89 Merge branch 4.x 2025-11-18 08:51:08 +03:00
Kumataro
6f74546488 Merge pull request #27318 from Kumataro:fix27298
core: add copyAt() for ROI operation #27318

Close https://github.com/opencv/opencv/issues/27320
Close https://github.com/opencv/opencv/issues/27298

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2025-11-17 10:58:20 +03:00
Alexander Smorkalov
64bbaa9f41 Merge pull request #27972 from asmorkalov:as/dst_mat_type
Force empty output type where it's defined in API #27972

The PR replaces:
- https://github.com/opencv/opencv/pull/27936
- https://github.com/opencv/opencv/pull/21059

Empty matrix has undefined type, so user code should not relay on the output type, if it's empty. The PR introduces some exceptions:
- copyTo documentation defines, that the method re-create output buffer and set it's type.
- convertTo has output type as parameter and output type is defined and expected.
2025-11-10 08:00:17 +03:00
Kumataro
879218500d Merge pull request #27918 from Kumataro:fix26899_5.x
core: support 16 bit LUT for 5.x #27918

Porting from https://github.com/opencv/opencv/pull/27890
Porting from https://github.com/opencv/opencv/pull/27911 

And support new OpenCV5 types for 16 bit LUT.

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
2025-10-20 08:30:28 +03:00