Robust Detection using sparse laser scanner with Autonomous Race Car

In this paper, a novel laser detection approach with lower objects is proposed. The detection method is based on RANSAC and another filter. For RANSAC, it detects ground surface and delete it in different driving condition. After that using Cluster, it can divide points into different groups which represents objects. Finally, there is a modified Formula SAE race car designed by author to finish experiment to test approach. We apply cones to combine track and do experiment in a random track to analysis the accuracy of detection.