基于改进欧几里得聚类的激光雷达障碍物检测

Obstacle Detection Based on Improved Euclidean Clustering with Lidar

针对传统的欧氏聚类算法无法同时对近处和远处的障碍物点云进行精准检测和分割,容易造成错检和漏检的情况,提出一种欧氏聚类算法的改进方法,可以根据点云与激光雷达之间的距离动态地选择阈值,从而快速且准确地完成聚类。实验表明:该方法能同时对近处和远处的障碍物点云进行快速且准确的聚类。

The traditional Euclidean clustering algorithm cannot accurately detect and segment the near and far obstacle point clouds at the same time,which is easy to cause false detection and missed detection.To solve the problems above,this paper proposes an improved method of Euclidean clustering algorithm,which can dynamically select the threshold according to the distance between the point cloud and the lidar,so as to complete the clustering quickly and accurately.Experiments show that this method can cluster the near and far obstacle point clouds quickly and accurately.