自動運転のための障害物死角領域の抽出

抄録

Recently, autonomous vehicle have been actively studied for reducing traffic accidents and conserving energy. It is necessary to recognize obstacles around the vehicle and determine drivable area. We use LIDAR for recognizing around the vehicle and make obstacle map that represent area ego vehicle is able to run. In the conventional map, we recognize area behind the nearest obstacle as unknown area. Therefore, we cannot get fine view in case of T-junction and intersection. So this paper deals with this problem to use obstacle height. We geometrically calculate blind area and make obstacle map suitable for actual road in order to reduce potential risk.