Moving Object Segmentation in 3D LiDAR Data: A Learning-based Approach Exploiting Sequential Data

Abstract

The ability to detect and segment moving objects in a scene is essential for building consistent maps, making future state predictions, avoiding collisions, and planning. In this paper, we address the problem of moving object segmentation from 3D LiDAR scans. We propose a novel approach that pushes the current state of the art in LiDAR-only moving object segmentation forward to provide relevant information for autonomous robots and other vehicles. Instead of segmenting the point cloud semantically, i.e., predicting the semantic classes such as vehicles, pedestrians, roads, etc., our approach accurately segments the scene into moving and static objects, i.e., also distinguishing between moving cars vs. parked cars. Our proposed approach exploits sequential range images from a rotating 3D LiDAR sensor as an intermediate representation combined with a convolutional neural network and runs faster than the frame rate of the sensor. We compare our approach to several other state-of-the-art methods showing superior segmentation quality in urban environments. Additionally, we created a new benchmark for LiDAR-based moving object segmentation based on SemanticKITTI. We published it to allow other researchers to compare their approaches transparently and we furthermore published our code.

arXiv commentAccepted by RA-L with IROS 2021

References in corpus (4)

• Open3D: A Modern Library for 3D Data Processing
• SuMa++: Efficient LiDAR-based Semantic SLAM
• Range Image-based LiDAR Localization for Autonomous Vehicles
• Multi-scale Interaction for Real-time LiDAR Data Segmentation on an Embedded Platform

Cited by in corpus (18)

• KISS-ICP: In Defense of Point-to-Point ICP -- Simple, Accurate, and Robust Registration If Done the Right Way
• Automatic Labeling to Generate Training Data for Online LiDAR-based Moving Object Segmentation
• Dynablox: Real-time Detection of Diverse Dynamic Objects in Complex Environments
• Meta-RangeSeg: LiDAR Sequence Semantic Segmentation Using Multiple Feature Aggregation
• Building Volumetric Beliefs for Dynamic Environments Exploiting Map-Based Moving Object Segmentation
• MotionBEV: Attention-Aware Online LiDAR Moving Object Segmentation with Bird's Eye View based Appearance and Motion Features
• Real-Time Metric-Semantic Mapping for Autonomous Navigation in Outdoor Environments
• Active Metric-Semantic Mapping by Multiple Aerial Robots
• Self-supervised Point Cloud Prediction Using 3D Spatio-temporal Convolutional Networks
• LiDAR Odometry Methodologies for Autonomous Driving: A Survey
• Radar Velocity Transformer: Single-scan Moving Object Segmentation in Noisy Radar Point Clouds
• RadarMOSEVE: A Spatial-Temporal Transformer Network for Radar-Only Moving Object Segmentation and Ego-Velocity Estimation
• 3D Adversarial Augmentations for Robust Out-of-Domain Predictions
• SemRaFiner: Panoptic Segmentation in Sparse and Noisy Radar Point Clouds
• Radar Tracker: Moving Instance Tracking in Sparse and Noisy Radar Point Clouds
• Weighted Bayesian Gaussian Mixture Model for Roadside LiDAR Object Detection
• Label-Free Model Failure Detection for Lidar-based Point Cloud Segmentation
• TwistSLAM++: Fusing multiple modalities for accurate dynamic semantic SLAM