In order to enhance the comprehensiveness of vehicle LiDAR sensor data collection,a method for filling missing data in new energy vehicle LiDAR sensors is proposed.The method utilizes data fusion-based point cloud acquisition techniques and median filtering for preprocessing the point cloud data.An improved noise density clustering algorithm is employed to construct point cloud supervoxels,build a graph model,and perform global clustering using graph cuts.By extracting typical features of objects and densely matching the missing regions using panoramic images,the missing areas in the point cloud data are filled.Experimental results demonstrate that the proposed method effectively fills missing data with good results.The filled point cloud data closely approximates the depth distribution of the original point cloud in the missing regions.
关键词
新能源汽车/激光雷达/传感器/缺失数据填补/点云采集/点云去噪
Key words
new energy vehicle/LiDAR/sensor/filling in missing data/point cloud collection/point cloud denoising
维普
万方数据