opencv_contrib is a repository of extra modules for OpenCV that extends the core library with contributed and experimental functionality.
The repository addresses the need to develop and test new computer vision modules without compromising the stability and binary compatibility of the main OpenCV distribution. New modules are developed here first, where they can evolve without the constraints of production-quality support. Once a module matures and gains adoption, it graduates to the central OpenCV repository where the development team provides full production support. This staging approach allows the community to contribute experimental features while keeping the core library reliable.
Developers should use opencv_contrib if they need cutting-edge computer vision functionality beyond what the stable OpenCV release offers, or if they are contributing new modules to the ecosystem. The modules are under constant development and work best alongside the master branch or latest releases of OpenCV. Individual modules can be selectively included or excluded during the build process using CMake options, so you need not compile everything if only specific functionality is required. The repository is designed for those comfortable with less-stable APIs and willing to track development branches rather than stable releases.
The project maintains active module development with ongoing contributions to the extra modules collection. Documentation is kept current through README files at both the repository and individual module levels, with clear guidance on module functionality, samples, and build requirements. The build integration with OpenCV core is actively supported through CMake configuration options and documented procedures for both command-line and GUI-based builds.
