校园物流车的激光雷达数据检测  

Lidar data detection of campus logistics vehicles

作　　者：何毅 袁锦 李冰林[1] 张涵 HE Yi;YUAN Jin;LI Binglin;ZHANG Han(College of Automobile and Traffic Engineering,Nanjing Forestry University,Nanjing 210037,Jiangsu,China)

机构地区：[1]南京林业大学汽车与交通工程学院,江苏南京210037

出　　处：《农业装备与车辆工程》2024年第2期74-78,共5页Agricultural Equipment & Vehicle Engineering

摘　　要：校园无人物流车能有效减轻快递点拥挤程度,提高配送效率。针对校园内人员通行无序、道路标记不明显等情况,采用Velodyne VLP-16激光雷达采集环境信息。对回波信号采用变分模态分解(VMD)的方法进行分解,再通过巴氏距离区分各模态间相关性,利用移动平均法处理非相关模态信号,再与相关模态信号进行重构。与其它滤波方法相比,该方法提高了信号的信噪比。通过RANSAC算法进行点云地面分割,再经欧式聚类算法进行校园道路障碍物点云聚类,实现了对障碍物的检测。Campus unmanned logistics vehicles can effectively reduce the congestion of express stores and improve delivery efficiency.In view of the disorderly passage of people in the campus,the road markings are not obvious,etc.,road environment information was collected useing the Velodyne VLP-16 Lidar.The echo signal was decomposed by the method of variational mode decomposition(VMD),and then the correlation between the modes was distinguished by the Bhattacharyian distance.The non-correlation mode was processed by moving average method,and then the correlation mode signal and the processed non-correlation mode signal were reconstructed to improve the signal-to-noise ratio of the signal compared with other filtering methods.The RANSAC algorithm was used to segment the point cloud ground,and then the European clustering algorithm was used to cluster the point cloud of campus road obstacles to realize the detection of obstacles.

关 键 词：校园物流车 激光雷达 VMD 点云分割 障碍物检测 

分 类 号：U463.54[机械工程—车辆工程]