A Study of Recognition Method of Curved Surface Reflector Code by LiDAR

In recent years, the convenience of road transportation has been remarkably improved by the spread of automobiles. However, there are problems of frequent traffic accidents and traffic congestions, and ITS (Intelligent Transport Systems) has been actively researched to solve these problems. ITS has several elements such as inter-vehicle communication, roadside-to-vehicle communication, and autonomous sensing with cameras and ladars. We propose a new autonomous sensing that uses infrared laser radar to detect reflector code on the roadside strip and acquire information. Infrared laser radar mounted on a vehicle scans the reflector code of the roadside strip while the vehicle is in motion. However, conventional methods have problems such as low speeds at which the vehicle can be scanned and decreased recognition accuracy as the vehicle speed increases. To solve these problems, this paper proposes to use LiDAR as a scanning device to improve vehicle speed and recognition accuracy. The intensity of the 3D point cloud data obtained by LiDAR can be used to discriminate the kind of reflectors. In the performance evaluation, we have confirmed that scanning at 30 km/h is possible. In the performance evaluation of recognition accuracy, we have confirmed that it can recognize the reflector code completely.