Vehicle-borne LiDAR point cloud extraction of vertical feature images in sector grid
For the problems of manual setting grid resolution and redundancy of storage resources by traditional method for road point extraction,a method for vehicle-borne LiDAR point cloud extraction of vertical feature images in sector grid is proposed in this paper.The method uses adaptive sector grid projection to automatically establish a scalable sector grid according to the parameters of LiDAR equipment,and effectively optimizes the manual resolution setting of traditional square grid.In the specific implementation of the method,the adaptive sector grid projection is firstly carried out according to the parameters of LiDAR equipment.Secondly,the elevation of each point in the grid is obtained,and the IDW method is used to establish the vertical feature images.Finally,the elevation threshold is set based on the vertical feature image and the road point is extracted.In order to illustrate the effectiveness of the proposed method,six sets of data of different road types are selected from the KITTI dataset for experiments,and the results showed that the best quality of road point extraction is 87.1％,the average quality is 80.2％,and the average quality of the proposed method is about three percentage higher than the traditional method.
real-time point cloudsadaptive sector gridIDWroad pointsKITTI dataset
NETL
NSTL
万方数据
