14 Pull requests merged by 10 people

• Added heuristic to allocate buffer for FAST features depending on input image resolution

  #27657 merged Aug 12, 2025

• Feat #25150: Pose Graph MST initialization

  #27423 merged Aug 12, 2025

• Bool masks support in imgproc module.

  #27644 merged Aug 11, 2025

• hal/riscv-rvv: implement FAST keypoint detection

  #27391 merged Aug 11, 2025

• test: gif: to test GIF codec at write_gif_flags

  #27649 merged Aug 11, 2025

• cuda 13.0 compatibility

  #27636 merged Aug 11, 2025

• Added Clip layer to new DNN engine

  #27624 merged Aug 8, 2025

• optimize some drawing with stack allocation

  #27599 merged Aug 7, 2025

• java: Set Automatic-Module-Name manifest attribute in the built JAR file

  #27577 merged Aug 7, 2025

• Update grammatical error in py_intro.markdown

  #27606 merged Aug 7, 2025

• Performance tests for writing and reading animations

  #27605 merged Aug 7, 2025

• Convert grayscale frames into color in interactive calibration preview

  #27628 merged Aug 7, 2025

• 5.x merge 4.x

  #27634 merged Aug 7, 2025

• Added options to initialize Orbbec cameras with custom fps and resolution

  #27629 merged Aug 6, 2025

13 Pull requests opened by 11 people

• core: fix missing vec_cvfo on IBM POWER9 due to unavailable VSX float64 conversion

  #27633 opened Aug 6, 2025

• fixed ocl.cpp data races without requiring additional locking

  #27638 opened Aug 6, 2025

• Try to fix KleidiCV on Linux and Mac.

  #27640 opened Aug 7, 2025

• Subdiv2d rect2f clean

  #27641 opened Aug 7, 2025

• stitching: enable loop unrolling in fast.cpp to improve ARM64 performance

  #27642 opened Aug 8, 2025

• fix Win32/Win64 garbled "Chinese" window titles in UNICODE builds. Fix #27645

  #27646 opened Aug 8, 2025

• Better reproducibility.

  #27647 opened Aug 8, 2025

• fix empty intrinsics handling and minor calibration improvements

  #27651 opened Aug 9, 2025

• Optimize VideoCapture with cvtColor instead of sws_scale

  #27652 opened Aug 10, 2025

• MGS-8318 Fix error:implicit declaration of function sub_group_reduce_add

  #27653 opened Aug 11, 2025

• Added proper type handling for TRILU layer

  #27655 opened Aug 11, 2025

• Added Size layer to new DNN engine

  #27656 opened Aug 11, 2025

• Added Determinant (Det) layer to new DNN engine

  #27658 opened Aug 11, 2025

6 Issues closed by 3 people

• Implement minimum spanning tree

  #25150 closed Aug 12, 2025

• cv::cuda::warpAffine does not work inplace

  #27429 closed Aug 11, 2025

• gif: Imgcodecs_Gif.write_gif_flags tests compression and decompression PNG image.

  #27648 closed Aug 11, 2025

• Link errors “undefined reference to cublasLtSetEnvironmentMode” & “cublasLtGetEnvironmentMode” when building OpenCV 4.13.0 with CUDA 12.9 on Ubuntu 24.04

  #27639 closed Aug 11, 2025

• Requesting support for Tensorflow 2 models

  #23427 closed Aug 7, 2025

• Crash in Faster-RCNN Object Detection

  #14839 closed Aug 7, 2025

8 Issues opened by 8 people

• Current TRILU layer implementation lacks proper type handling

  #27654 opened Aug 11, 2025

• multiview_calibration.py fails when --intrinsics contains empty entries (cv2.FileStorage("") error)

  #27650 opened Aug 9, 2025

• Win32/Win64 imshow() always shows window with garbled "Chinese" titles (UNICODE build)

  #27645 opened Aug 8, 2025

• OpenCV to MSMF

  #27643 opened Aug 8, 2025

• Data races in ocl.cpp

  #27637 opened Aug 6, 2025

• Missing vec_cvfo symbol on POWER9 when building downstream project with OpenCV 4.x/3.x

  #27635 opened Aug 6, 2025

• dml_ep.obj : error LNK2019: unresolved external symbol

  #27632 opened Aug 6, 2025

• Unable to build `opencv_highgui` with Qt 6.9.1

  #27630 opened Aug 5, 2025

36 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• modules/videoio: add libcamera backend for OpenCV VideoCapture

  #27554 commented on Aug 12, 2025 • 11 new comments

• Trilu fix

  #27619 commented on Aug 11, 2025 • 3 new comments

• Add Alpha matting samples (C++ and Python)

  #27593 commented on Aug 12, 2025 • 3 new comments

• Add seemore_superres sample for image super-resolution using seemoredetails (C++ and Python)

  #27592 commented on Aug 12, 2025 • 3 new comments

• Fix issues in Convex Hull algorithm

  #27020 commented on Aug 7, 2025 • 2 new comments

• core: ARM64 loop unrolling in kmeans to improve Weighted Filter performance

  #27596 commented on Aug 11, 2025 • 1 new comment

• Added fully functional resize layer to new DNN engine

  #27586 commented on Aug 11, 2025 • 1 new comment

• OpenCV upgrade 4.11.0 -> 4.12.0: Release and Debug version overwrite each other upon installation

  #27626 commented on Aug 5, 2025 • 0 new comments

• [GSOC 2025] Imgcodecs Metadata Support

  #27488 commented on Aug 7, 2025 • 0 new comments

• [WIP] GSoC 2025: Add Tokenizer Support to DNN Module

  #27534 commented on Aug 8, 2025 • 0 new comments

• Refactor Blackwell

  #27537 commented on Aug 8, 2025 • 0 new comments

• imgcodecs: bmp: support 32bpp BMP with IMREAD_UNCHANGED

  #27559 commented on Aug 8, 2025 • 0 new comments

• Added fully functional convTranspose layer to new DNN engine

  #27560 commented on Aug 11, 2025 • 0 new comments

• dlpack Python support for cuda_GpuMat and cuda_GpuMatND

  #27581 commented on Aug 7, 2025 • 0 new comments

• G-API: Implement cfgClampOutputs option to OpenVINO Params

  #27600 commented on Aug 11, 2025 • 0 new comments

• Support Symbol.dispose for using keyword in OpenCV.js (#26157)

  #27613 commented on Aug 7, 2025 • 0 new comments

• Fix Python Scalar typing issue #27528

  #27620 commented on Aug 12, 2025 • 0 new comments

• Add strict validation for encoding parameters

  #27621 commented on Aug 11, 2025 • 0 new comments

• add a muti-camera visualisation script

  #27627 commented on Aug 7, 2025 • 0 new comments

• Incorrect floating point edge case handling in cv::Mat::convertTo() and cv::saturate_cast() due to cvRound()

  #27556 commented on Aug 5, 2025 • 0 new comments

• CAP_PROP_WHITE_BALANCE_BLUE_U incorrectly controls white balance temperature on Windows (DirectShow backend)

  #27372 commented on Aug 6, 2025 • 0 new comments

• cuda::remap() returns different results from cv::remap()

  #24602 commented on Aug 6, 2025 • 0 new comments

• HDR tutorial result is different from the code running.

  #22219 commented on Aug 7, 2025 • 0 new comments

• Why Subdiv2D takes ROI rect as cv::Rect and not cv::Rect2f? (possible bug)

  #27623 commented on Aug 7, 2025 • 0 new comments

• OpenCV 5: use `expected` for cleaner APIs

  #27272 commented on Aug 7, 2025 • 0 new comments

• Any intention to add interoperability with DLPACK or mpi4py?

  #16295 commented on Aug 7, 2025 • 0 new comments

• std::stringview usage on imread/write

  #27562 commented on Aug 8, 2025 • 0 new comments

• Link OpenCV with Cmake FetchContent

  #20548 commented on Aug 9, 2025 • 0 new comments

• Add support for libcamera

  #21653 commented on Aug 9, 2025 • 0 new comments

• LLM integration prototype

  #27176 commented on Aug 10, 2025 • 0 new comments

• opencv_perf_video meet implicit declaration of function 'sub_group_reduce_add' is invalid in OpenCL error

  #27507 commented on Aug 11, 2025 • 0 new comments

• After setting the decoder to hevc_cuvid, VideoCapture reads video more slowly

  #21969 commented on Aug 11, 2025 • 0 new comments

• OpenCV4.8.0 DNN inference speed reduced by 50%.

  #23911 commented on Aug 12, 2025 • 0 new comments

• 4.12.0 crash with android targetSdk 35 16KB page size

  #27618 commented on Aug 12, 2025 • 0 new comments

• [WIP] modules/videoio: add libcamera backend for OpenCV Video I/O

  #27155 commented on Aug 9, 2025 • 0 new comments

• G-API: Add support to set workload type dynamically in both OpenVINO and ONNX OVEP

  #27460 commented on Aug 11, 2025 • 0 new comments