An approach to robust ICP initialization
From MaRDI portal
Publication:6420146
arXiv2212.05332MaRDI QIDQ6420146FDOQ6420146
Authors: Alexander Kolpakov, Michael Werman 
Publication date: 10 December 2022
Abstract: In this note, we propose an approach to initialize the Iterative Closest Point (ICP) algorithm to match unlabelled point clouds related by rigid transformations. The method is based on matching the ellipsoids defined by the points' covariance matrices and then testing the various principal half-axes matchings that differ by elements of a finite reflection group. We derive bounds on the robustness of our approach to noise and numerical experiments confirm our theoretical findings.
Has companion code repository: https://github.com/sashakolpakov/icp-init
This page was built for publication: An approach to robust ICP initialization
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6420146)
