About Wasmvision
wasmVision is an open source computer vision processing engine. It includes vision algorithms and machine learning models, and it is built to be customized and extended with WebAssembly, so new processing steps can be added without changing the core engine.
A pipeline has three stages. The engine captures input from cameras, images, video files or streams, processes the video with computer vision algorithms and ML models, and sends the results to a stream, an image, a video file or a data store. Built in processors cover common tasks such as face detection and edge detection, and developers can write their own processors in Go, Rust or C. The engine can run on embedded devices or in the cloud and takes advantage of hardware acceleration.
It is built on established open source projects, including OpenCV, Wazero and GoCV. The audience is developers and engineers who want to prototype or deploy video analysis pipelines, particularly where a lightweight, portable runtime matters, such as edge hardware.
The site links to a getting started guide and to the source code. wasmVision is free open source software, and the site publishes no paid plans or pricing.