PGP-DOR: A Point-Grid-Point Scheme for Efficient Dynamic Object Removal
Shuo Wang, Zhenping Sun, Hanzhang Xue, Bokai Liu, Hao Fu, YinFu Luo
AI summary
Problem
Dynamic objects in LiDAR scans introduce noise and artifacts into high-precision maps, degrading autonomous navigation. Existing methods either rely on costly annotated datasets, lack generalizability across LiDAR types, or suffer from high computational costs and false positives in complex scenes.
Approach
The method estimates dynamic attributes at the 3D point level, aggregates them into a 2D bird’s-eye view grid for dense spatio-temporal inference, and refines the results back to the point level to accurately classify moving versus static objects.
Key results
- State-of-the-art accuracy on SemanticKITTI and a custom multi-LiDAR dataset
- Robust performance in both online and offline mapping settings
- Effective suppression of ground false positives and handling of LiDAR sparsity
- Balanced computational efficiency with high precision compared to baselines
Why it matters
Enables reliable, high-precision map construction for autonomous vehicles without dependency on large-scale annotated datasets or specific LiDAR hardware.
Abstract
In the field of autonomous driving, constructing high-precision maps, typically represented as 3D point cloud maps or bird’s-eye view (BEV) grid maps, is essential for both offline and online applications. However, the presence of dynamic objects within a scene can introduce artifacts and noise that significantly degrade the quality of these maps. To address this challenge, we propose a method in this paper that can accurately identify those dynamic objects in both online and offline settings. Our approach fully exploits the spatio-temporal attributes of BEV grid maps and utilizes a point-grid-point (PGP) scheme to identify moving objects at both the 3D point cloud level and the 2D BEV grid level. Experimental results from public datasets, as well as a self-collected dataset, demonstrate that our method consistently outperforms state-of-the-art approaches in dynamic object removal in both online and offline contexts. The code and the newly introduced dataset will be made publicly available at: https://github.com/MichealRW/PGP-DOR .