SuperLine3D: Self-supervised Line Segmentation and Description for LiDAR Point Cloud

TL;DR

This paper introduces SuperLine3D, a novel self-supervised learning model for 3D line segmentation and description in LiDAR point clouds, enabling robust urban scene analysis and registration without extensive manual labeling.

Contribution

It presents the first learning-based approach for 3D line segmentation and description in LiDAR data, using synthetic data and iterative auto-labeling for training.

Findings

Effective line segmentation under scale variations

Competitive registration accuracy compared to point-based methods

Self-supervised training reduces labeling effort

Abstract

Poles and building edges are frequently observable objects on urban roads, conveying reliable hints for various computer vision tasks. To repetitively extract them as features and perform association between discrete LiDAR frames for registration, we propose the first learning-based feature segmentation and description model for 3D lines in LiDAR point cloud. To train our model without the time consuming and tedious data labeling process, we first generate synthetic primitives for the basic appearance of target lines, and build an iterative line auto-labeling process to gradually refine line labels on real LiDAR scans. Our segmentation model can extract lines under arbitrary scale perturbations, and we use shared EdgeConv encoder layers to train the two segmentation and descriptor heads jointly. Base on the model, we can build a highly-available global registration module for point cloud registration, in conditions without initial transformation hints. Experiments have demonstrated that our line-based registration method is highly competitive to state-of-the-art point-based approaches. Our code is available at https://github.com/zxrzju/SuperLine3D.git.