激光与光电子学进展, 2019, 56 (24): 242801, 网络出版: 2019-11-26
基于三维激光雷达的障碍物及可通行区域实时检测 下载: 1706次
Real-Time Detection of Obstacles and Passable Areas Based on Three-Dimensional Lidar
遥感 三维激光雷达 障碍物检测 改进欧氏聚类算法 可通行区域 相邻点云间距算法 remote sensing three-dimensional lidar obstacle detection improved Euclidean clustering algorithm passable area adjacent point cloud spacing algorithm
摘要
针对交通环境中障碍物及可通行区域检测的问题,利用改进欧氏聚类算法进行实时障碍物检测,提出一种相邻点云间距算法以实时提取道路可通行区域。对点云数据进行预处理,再通过地面坡度分离算法进行地面与非地面点云分离;根据不同的聚类距离阈值对非地面点云进行障碍物聚类检测,使用长方体框标记区分;将每个地面激光束固有的相邻点云间距与实际相邻两点间距离进行对比,结合相邻点角度差以及点云聚类,实现可通行区域的提取;融合障碍物检测和可通行区域提取结果,对可通行区域的通过性进行合并检测。多路况实车实验表明,本文算法能准确检测出障碍物及道路的可通行区域,检测平均准确率为94.13%,平均耗时为69 ms,满足智能车实时性的要求。
Abstract
of obstacle detection and passable area extraction are fused to detect the passing ability of the passable area. Multi-road real-time vehicle experiments show that the proposed algorithm can accurately detect obstacles and passable road areas, with an average detection accuracy of 94.13% and an average processing time of 69 ms, meeting the real-time requirements of intelligent vehicles.
蒋剑飞, 李其仲, 黄妙华, 龚杰. 基于三维激光雷达的障碍物及可通行区域实时检测[J]. 激光与光电子学进展, 2019, 56(24): 242801. Jianfei Jiang, Qizhong Li, Miaohua Huang, Jie Gong. Real-Time Detection of Obstacles and Passable Areas Based on Three-Dimensional Lidar[J]. Laser & Optoelectronics Progress, 2019, 56(24): 242801.