On June 3, the latest release of OpenCV version 4.10 was announced. It features numerous improvements in the Deep Neural Network (DNN) module, experimental support for NumPy 2.0, various enhancements for Android, support for NVIDIA CUDA 12.4+, improvements for RISC-V and ARM, support for oneAPI 2024, experimental support for Windows ARM64, experimental support for Apple VisionOS, and built-in Wayland server support in Linux. Various issues regarding Wayland support in the OpenCV HighGUI code have been resolved.
OpenCV (or Open Source Computer Vision Library) is a library of computer vision algorithms that support machine learning, object detection, segmentation and recognition, motion tracking in video, gesture recognition, and many other functions and related tasks. The library was originally implemented in C, and since the second version, it has been transitioning to C++. Bindings for Python, Java, Ruby, MATLAB/Octave, Lua, and other languages are supported. It can be used freely for academic and commercial purposes under the BSD license. A significant part of the project was developed by the Russian company Itseez, which was acquired by Intel in May 2016.
Supported OS:
- Microsoft Windows (compilers Microsoft Visual C++ (6.0, .NET 2003), Intel Compiler, Borland C++, Mingw (GCC 3.x))
- Microsoft Windows RT
- Linux (GCC 2.9x, 3.x, Intel Compiler)
- macOS
- Android
- iOS
Support is provided for graphical interface and video capture tools in Windows — DirectShow, VfW, MIL, CMU1394; in Linux — V4L2, DC1394, FFMPEG; in macOS — QuickTime.
Source: linux.org.ru
