In This paper, we present a method for obstacles detection in sparse point cloud from Lidar. For point cloud collected by rotative mechanical lidar, we firstly segment point cloud into ground points and non-ground points. Inspired by the former scholars’ work, we proposed a hybrid model based ground filtering method that combines simple line model and Gaussian model regression, which achieve real-time efficiency and strong ground representation ability for undulating ground environment. And in point cloud clustering process, we proposed a multi-dimensional point similarity measuring principle, then construct a KD-Tree to accelerate the clustering process. Compared with popular methods and deep learning method, ours work solved the obstacles detection more efficient with competitive accuracy abstract should summarize the contents of the paper in short terms.
An Obstacles Detection Method for Sparse PointCloud Based on Hybrid Model and Multi-dimensional Segmentation
Lect. Notes Electrical Eng.
International Conference on Intelligent Transportation Engineering ; 2021 ; Beijing, China October 29, 2021 - October 31, 2021
2021 6th International Conference on Intelligent Transportation Engineering (ICITE 2021) ; Kapitel : 106 ; 1198-1209
01.06.2022
12 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Virtual validation of autonomous vehicles in a 3D Pointcloud
British Library Conference Proceedings | 2018
|