Efficient variants of the ICP algorithm

TLDR

ICP is a widely used algorithm for aligning 3D models given an initial pose, and numerous variants have been developed that modify point selection, matching, and minimization. The study catalogs and classifies ICP variants, evaluates their impact on convergence speed, introduces a uniform-normal-sampling variant for nearly-flat meshes, and proposes a high-speed optimized combination. They systematically enumerate and classify existing ICP variants, assess their convergence speed experimentally, develop a new uniform-normal-sampling variant for flat meshes, and combine selected variants to achieve high speed. The implementation aligns two range images in a few tens of milliseconds given a good initial guess, demonstrating potential for real-time 3D acquisition and model-based tracking.

Abstract

The ICP (Iterative Closest Point) algorithm is widely used for geometric alignment of three-dimensional models when an initial estimate of the relative pose is known. Many variants of ICP have been proposed, affecting all phases of the algorithm from the selection and matching of points to the minimization strategy. We enumerate and classify many of these variants, and evaluate their effect on the speed with which the correct alignment is reached. In order to improve convergence for nearly-flat meshes with small features, such as inscribed surfaces, we introduce a new variant based on uniform sampling of the space of normals. We conclude by proposing a combination of ICP variants optimized for high speed. We demonstrate an implementation that is able to align two range images in a few tens of milliseconds, assuming a good initial guess. This capability has potential application to real-time 3D model acquisition and model-based tracking.