Research on Automatic Extraction Method of Traffic Marking Based on Vehicle LiDAR
In view of the increasing demand of autonomous driving technology for real-time storage and analysis of high-precision road information and the redundant distization of road point cloud data, this paper proposes an effective method to automatically extract road surface, classify and vectorize traffic lines from on-board LiDAR point cloud data. Firstly, non-ground points in point cloud data are filtered out; secondly, pseudo-scanning lines are generated based on the track lines of carrier vehicles to realize road surface extrac-tion; then, a series of 2D point cloud reference images are constructed, and the point cloud intensity and other characteristic informa-tion are used to detect the boundary pixels and coordinates of traffic markings, and the outliers are removed to classify and refine traffic markings. Finally, the traffic elements extracted by the proposed method are compared with those extracted by the traditional method. The experimental results show that the accuracy and efficiency of the proposed method have been improved.
vehicle LiDARroad surfacetraffic markingpoint cloud features
万方数据
维普
