IRON: Invariant-based Highly Robust Point Cloud Registration

TL;DR

IRON is a robust point cloud registration method that effectively handles up to 99% outliers by decoupling the problem, estimating scale, rotation, and translation separately, and employing invariant compatibility and convex relaxation techniques.

Contribution

The paper introduces IRON, a novel registration algorithm combining invariant-based inlier detection, convex SDP relaxation, and an outlier rejection heuristic, improving robustness and efficiency.

Findings

Outperforms state-of-the-art algorithms in accuracy and robustness.

Handles up to 99% outliers effectively.

Demonstrates efficiency and high accuracy on real datasets.

Abstract

In this paper, we present IRON (Invariant-based global Robust estimation and OptimizatioN), a non-minimal and highly robust solution for point cloud registration with a great number of outliers among the correspondences. To realize this, we decouple the registration problem into the estimation of scale, rotation and translation, respectively. Our first contribution is to propose RANSIC (RANdom Samples with Invariant Compatibility), which employs the invariant compatibility to seek inliers from random samples and robustly estimates the scale between two sets of point clouds in the meantime. Once the scale is estimated, our second contribution is to relax the non-convex global registration problem into a convex Semi-Definite Program (SDP) in a certifiable way using Sum-of-Squares (SOS) Relaxation and show that the relaxation is tight. For robust estimation, we further propose RT-GNC (Rough Trimming and Graduated Non-Convexity), a global outlier rejection heuristic having better robustness and time-efficiency than traditional GNC, as our third contribution. With these contributions, we can render our registration algorithm, IRON. Through experiments over real datasets, we show that IRON is efficient, highly accurate and robust against as many as 99% outliers whether the scale is known or unknown, outperforming the existing state-of-the-art algorithms.