RMBP | Robust Matching using Belief Propagation | GPU library

RMBP is a C++ library typically used in Hardware, GPU, Pytorch applications. RMBP has no bugs, it has no vulnerabilities, it has a Permissive License and it has low support. You can download it from GitHub.

RMBP is a robust matching approach, aiming at rejecting outlier 2D or 3D keypoint matches based on the inference via belief propagation on the defined graphical model. Therefore, it can contribute to the robustness of point cloud registration and image matching. As shown below, RMBP facilitates the registration of point clouds with deficient geometric overlap although the outlier matches outnumber the inlier matches greatly in such cases. We enable the application of RMBP in two scenarios in this repository: 3D-3D matching and 2D-2D matching, which correspond to the matster and 2D-2D branches respectively.