为有效检测道路中的障碍物,识别其中的车辆目标,提出一种基于三维激光雷达的障碍物检测和车辆识别方法。采用雷达相邻扫描点的距离特征分割地面,利用DBSCAN聚类方法对障碍物聚类,利用IEPF算法分割和拟合障碍物轮廓曲线,最后结合多个判据实现车辆目标检测。在城市快速路进行实验,结果表明该方法可以有效识别其中的车辆目标。
To effectively detect obstacles in road and identify vehicle targets from those obstacles, the paper puts forward a method detecting obstacles and vehicle based on 3D lidar. Firstly, it divides ground with neighbor scanning point of lidar and gathers obstacles with DBSCAN clustering method. Then, it divides and fits profile curve of obstacles with IEPF algorithm. Finally, it realizes vehicle targets detection with multiple criterions. The experiment on urban expressway shows that this method can effectively identify the vehicle targets.