Lane Recognition Using On-vehicle LIDAR

This paper presents an approach to the lane recognition using on-vehicle LIDAR. It detests the objects by 2D scanning and collects the range and reflectivity data in each scanning direction. We developed the lane recognition algorithm with these data, in which the lane curvature, yaw angle and offset are calculated by using the Hough transformation, and the lane width is calculated by statistical procedure. Next the lane marks are tracked by the extended Kalman filter. Then we test the performance of the lane recognition and the good results are achieved. Finally, we show the result of the road environment recognition applying the lane recognition by LIDAR