Research Article
Low-Cost LiDAR-Based Vehicle Detection for Self-driving Container Trucks at Seaport
@INPROCEEDINGS{10.1007/978-3-030-92638-0_27, author={Changjie Zhang and Zhenchao Ouyang and Lu Ren and Yu Liu}, title={Low-Cost LiDAR-Based Vehicle Detection for Self-driving Container Trucks at Seaport}, proceedings={Collaborative Computing: Networking, Applications and Worksharing. 17th EAI International Conference, CollaborateCom 2021, Virtual Event, October 16-18, 2021, Proceedings, Part II}, proceedings_a={COLLABORATECOM PART 2}, year={2022}, month={1}, keywords={Obstacle detection and tracking Autonomous truck Multiple LiDAR Deep learning Seaport area}, doi={10.1007/978-3-030-92638-0_27} }- Changjie Zhang
Zhenchao Ouyang
Lu Ren
Yu Liu
Year: 2022
Low-Cost LiDAR-Based Vehicle Detection for Self-driving Container Trucks at Seaport
COLLABORATECOM PART 2
Springer
DOI: 10.1007/978-3-030-92638-0_27
Abstract
The self-driving technology has been developed rapidly in the past decades, due to new sensors, and car manufacturers have become more open. However, fully self-driving vehicles for the public still has a long way to go. Most studies try to focus on self-driving in special scenes, such as park sightseeing car, express logistics vehicle, sweeper, indoor service robot, and special vehicles in the mining area or seaport area. One of the critical issues is that the cost of a self-driving vehicle should strictly be controlled for commercial uses. This paper presents a low-cost LiDAR-based moving obstacle detection and tracking for self-driving container trucks in the low-speed seaport area. We build a CNN model for obstacle detection with the bird’s eye view (BEV) map generated from two low density LiDARs equipped at the head of a container truck. A boosting tracker is used to achieve real-time processing speed on the embedded module of Tx2. Simulation on the collected data shows that our Strided-Yolo model can achieve the highest mAP on the BEV projection map than other models.
