In order to solve the problem that the traffic-cone detection accuracy of Formula Student unmanned vehicles in the track environment is not high,this paper proposes a three-dimensional traffic-cone target detection algorithm for Formula Student unmanned vehicle competition.The algorithm first extracts the ROI area from the point cloud collected by the lidar,then removes the ground points in the area,and finally performs point cloud clustering in the non-ground point cloud,and clusters the points belonging to the traffic-cone into a cluster to realize the detection of the traffic-cone in the track.This method improves the original single-attribute clustering method,and uses multiple attributes of point cloud intensity and density for clustering.Through the actual vehicle test of Formula Student unmanned vehicle,the three-dimensional traffic-cone object detection algorithm proposed in this paper achieves more than 90%accuracy in multiple track scenes,providing excellent performance algorithm for subsequent Formula Student unmanned vehicle competition.
维普
万方数据
