基于激光雷达的道路可通行区域检测

Road Available Area Detecting Based on Laser Scanner

随着城市化进程的发展,城市人口密度日益增高,随之而来的城市交通拥堵问题也日渐显著,而无人驾驶车辆作为改善这一问题的一个可能手段,也得到了越来越多的关注。无人驾驶车辆的安全性无疑是其研发过程中十分重要的一环。该文提出了一种使用3个激光雷达进行的道路可通行区域检测方法,该方法包括道路边界检测与障碍物检测两部分。道路边界检测将道路剖面的弧度考虑在内,使用了二次函数进行拟合,提高了模型的准确性,并将最终结果表示于1个以圆弧为基础的栅格地图内。该表示方法能更好地贴合车辆实际运动模型。

As the continuously growing city population, the higher and higher density of vehicles has become a significant issue. The intelligent vehicles may be a solution against it. Safety is a basic requirement, both for the passengers and the pedestrians. This paper proposes a road available area detecting method with three laser scanners, two of which are used to detect the bound of the road, and the rest one is for obstacle detection. The result is expressed in a circle-based raster map, which fit the dynamic model of car movement better.