A comprehensive survey on point cloud registration

TL;DR

This survey comprehensively reviews point cloud registration methods, highlighting recent advances, connections between optimization and deep learning approaches, and introduces a new benchmark for cross-source registration challenges.

Contribution

It provides an extensive overview of both same-source and cross-source registration methods, clarifies their relationships, and establishes a new benchmark dataset for evaluating algorithms.

Findings

Deep learning and optimization methods have improved robustness and efficiency.

Connections between optimization-based and deep learning methods are still unclear.

A new benchmark dataset for cross-source point cloud registration is introduced.

Abstract

Registration is a transformation estimation problem between two point clouds, which has a unique and critical role in numerous computer vision applications. The developments of optimization-based methods and deep learning methods have improved registration robustness and efficiency. Recently, the combinations of optimization-based and deep learning methods have further improved performance. However, the connections between optimization-based and deep learning methods are still unclear. Moreover, with the recent development of 3D sensors and 3D reconstruction techniques, a new research direction emerges to align cross-source point clouds. This survey conducts a comprehensive survey, including both same-source and cross-source registration methods, and summarize the connections between optimization-based and deep learning methods, to provide further research insight. This survey also builds a new benchmark to evaluate the state-of-the-art registration algorithms in solving cross-source challenges. Besides, this survey summarizes the benchmark data sets and discusses point cloud registration applications across various domains. Finally, this survey proposes potential research directions in this rapidly growing field.