Abstract:Three-dimensional point cloud registration is an important field in computer vision. Recently, due to the increasingly complex scenes and incomplete observations, many partial-overlap registration methods based on overlap estimation have been proposed. These methods heavily rely on the extracted overlapping regions with their performances greatly degraded when the overlapping region extraction underperforms. To solve this problem, we propose a partial-to-partial registration network (RORNet) to find reliable overlapping representations from the partially overlapping point clouds and use these representations for registration. The idea is to select a small number of key points called reliable overlapping representations from the estimated overlapping points, reducing the side effect of overlap estimation errors on registration. Although it may filter out some inliers, the inclusion of outliers has a much bigger influence than the omission of inliers on the registration task. The RORNet is composed of overlapping points' estimation module and representations' generation module. Different from the previous methods of direct registration after extraction of overlapping areas, RORNet adds the step of extracting reliable representations before registration, where the proposed similarity matrix downsampling method is used to filter out the points with low similarity and retain reliable representations, and thus reduce the side effects of overlap estimation errors on the registration. Besides, compared with previous similarity-based and score-based overlap estimation methods, we use the dual-branch structure to combine the benefits of both, which is less sensitive to noise. We perform overlap estimation experiments and registration experiments on the ModelNet40 dataset, outdoor large scene dataset KITTI, and natural data Stanford Bunny dataset. The experimental results demonstrate that our method is superior to other partial registration methods. Our code is available at https://github.com/superYuezhang/RORNet.

OKR-Net: Overlapping Keypoints Registration Network for Large-Scale LiDAR Point Clouds

RORNet: Partial-to-Partial Registration Network With Reliable Overlapping Representations

Point Cloud Registration using Representative Overlapping Points

Efficient low-overlapping point cloud registration based on two-stage network learning

MSPR-Net: A Multi-Scale Features Based Point Cloud Registration Network

Sparse and Low-Overlapping Point Cloud Registration Network for Indoor Building Environments

Robust Point Cloud Registration Network for Complex Conditions

HRegNet: A Hierarchical Network for Large-scale Outdoor LiDAR Point Cloud Registration

Efficient and Robust Line-based Registration Algorithm for Robot Perception under Large-scale Structural Scenes

HRegNet: A Hierarchical Network for Efficient and Accurate Outdoor LiDAR Point Cloud Registration

RRGA-Net: Robust Point Cloud Registration Based on Graph Convolutional Attention

OMNet: Learning Overlapping Mask for Partial-to-Partial Point Cloud Registration

Iterative Overlap Attention-Aware Network with Similarity Learning for Partial Point Cloud Registration

JoKDNet: A Joint Keypoint Detection and Description Network for Large-Scale Outdoor TLS Point Clouds Registration.

OPSNet: Point Cloud Registration Based on Overlapping Predictive Segmentation

Robust Point Cloud Registration Using Geometric Spatial Refinement

Rethinking of learning-based 3D keypoints detection for large-scale point clouds registration

Pyramid Semantic Graph-based Global Point Cloud Registration with Low Overlap

PointNetLK: Robust & Efficient Point Cloud Registration using PointNet

DDRNet: Fast point cloud registration network for large-scale scenes

A Novel and Efficient Point Cloud Registration by Using Coarse-to-Fine Strategy Integrating PointNet