cv

Why we included this project

Developers doing geometry-heavy computer vision in Rust have a genuine alternative to binding into OpenCV. The mono-repo gathers pure-Rust crates covering camera models, feature detection and description, geometric estimation, multi-view triangulation, and structure-from-motion. That covers photogrammetry, pose estimation, SLAM-style pipelines, and 3D reconstruction. Each crate is an independent dependency with its own documentation, so you can pull in just the eight-point algorithm or the Akaze feature detector instead of adopting a whole framework. The tutorials and the batteries-included `cv` crate help get a first pipeline running quickly. This is not a deep-learning library; it targets the traditional algorithmic layer rather than neural networks and GPU inference.