Point Cloud Segmentation and Detection for Vehicle Based on LIDAR Sensor
Download as PDF
Xuan Xiu, Xianglei Zhu, Lianqing You
Currently automatic driving technology is the most rapid research direction in the field of artificial intelligence. Besides, the environment sensor LIDAR is the most significant hardware in system automatic sensor system. This paper concentrates on the cloud data acquired by the LIDAR as the input data and utilizes the point cloud library algorithm to write C++ program. We not only achieve the single vehicle object segmentation and detection in open scenario but realize the same target of the multiple vehicles in parking scene. We obtain a comparatively nice result of vehicle’s segmentation and detection. Simultaneously, our C++ program package provides the foundation for further research on point cloud recognition and semantic level segmentation.
Intelligent driving, Velodyne LIDAR, 3D point cloud segmentation, vehicle point cloud detection, random sample consensus method, Euclidean segmentation, Hough clustering