A LiDAR-based driving perception system is crucial for autonomous driving, as it provides real-time, accurate, and robust perception results that significantly enhance driving safety. This article proposes a LiDAR-based panoptic driving perception algorithm, named LiDAR-PDP, which can simultaneously perform free-space segmentation, static object semantic segmentation, and moving object instance segmentation. This article constructs an efficient two-branch 2-D convolution neural network (CNN) backbone that consists of a spatial feature branch and a motion feature branch to extract fine-grained spatial-temporal features from sequential point cloud data. For the spatial feature branch, we designed a compound sampling convolution module (CSCM) that can learn 3-D spatial features with a more general geometric modeling capacity. Additionally, this article designs parallel feature pyramid networks (P-FPNs) that can aggregate multiscale features in both range-dependent and size-dependent manners. We evaluate LiDAR-PDP on SemanticKITTI and NuScenes. Our method outperforms other comparison methods that were modified from the state-of-the-art point cloud panoptic segmentation methods on both datasets, thereby demonstrating the effectiveness of the proposed components in LiDAR-PDP. Moreover, our lightweight LiDAR-PDP achieves outstanding accuracy and real-time performance on a vehicular and edge computing platform. The LiDAR-PDP would improve the perception ability of autonomous vehicles with limited computing resources.
LiDAR-PDP: A LiDAR-Based Panoptic Dynamic Driving Environment Perception Algorithm
IEEE Transactions on Transportation Electrification ; 11 , 1 ; 1848-1862
01.02.2025
5854705 byte
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Panoptic Based Camera and Lidar Fusion for Distance Estimation in Autonomous Driving Vehicles
SAE Technical Papers | 2022
|Panoptic Based Camera and Lidar Fusion for Distance Estimation in Autonomous Driving Vehicles
British Library Conference Proceedings | 2022
|Panoptic Based Camera and Lidar Fusion for Distance Estimation in Autonomous Driving Vehicles
British Library Conference Proceedings | 2022
|MtpNet: Multi-Task Panoptic Driving Perception Network
IEEE | 2025
|