基于二维LiDAR的轨道侵入异物检测系统与方法研究  

Research on track invasion foreign object detection system and method based on two-dimensional LiDAR

作　　者：李志南 刘玮 韩丽娟[2] 殷思强 LI Zhi-nan;LIU wei;HAN Li-juan;YIN Si-qiang(Locomotive Department of Hebei Rail Transportation Vocational and Technical College,Hebei Shijiazhuang 050000,China;College of Mechanical and Electrical Engineering,Hebei Agricultural University.Hebei Baoding 071001,China;Beijing Key Laboratory of Electronic Engineering Technology,Tsinghua University,Beijing 100084,China)

机构地区：[1]河北轨道运输职业技术学院机车系,河北石家庄050000 [2]河北农业大学机电工程学院,河北保定071001 [3]清华大学光电子工程与技术北京市重点实验室,北京100084

出　　处：《光电子．激光》2020年第3期262-268,共7页Journal of Optoelectronics·Laser

基　　金：河北省教育科学“十三五”规划课题(1803090)资助项目。

摘　　要：为了实现对铁路侵入异物的检测,并克服摄像头检测可靠性低以及激光雷达检测数据单一的缺点。设计了一种融合摄像头、二维激光雷达的异物目标体检测系统,将软件测试结果可视化。在二维激光雷达的基础上,建立基于凝聚层次聚类算法的目标体识别方法,对单个人以及三个人开展了目标识别实验,并同时建立了基于限幅滤波法的室内、实地的单目标体与多目标体追踪方法,对多目标体在追踪丢失、重合、遮盖情况下,开展了目标追踪实验。最终,将二维激光雷达与摄像头进行信息融合,分别对单目标体以及多目标体进行检测实验。结果显示:激光雷达所采集点云数据经过凝聚层次聚类算法以及去噪处理可以较好的实现对目标体的识别,限幅滤波法可以对追踪目标体短时间内发生丢失、被遮挡或与其他目标体重合的情况下,再次进行追踪,二维激光雷达与摄像头信息融合的结果可以较好的得到不同目标体的位置与边界。该研究结果可为铁路安全检测系统的智能化提供理论参考与依据。In order to achieve the detection of railway intrusion into foreign bodies,and overcome the shortcomings of low reliability of camera detection and single detection data of lidar.A foreign object detection system combining camera and 2 D radar was designed to visualize the software test results.Based on the two-dimensional lidar,the target recognition method based on condensed hierarchical clustering algorithm is established.The target recognition experiment is carried out for a single person and three people.At the same time,the indoor and field single targets based on Kalman filtering are established.The body and multi-target tracking method is used to perform target tracking experiments on multi-targets in the case of tracking loss,coincidence and occlusion.Finally,the two-dimensional lidar and the camera are combined with information to test the single target and multiple targets.The results show that the point cloud data collected by the laser radar can realize the recognition of the target body through the condensed hierarchical clustering algorithm and the denoising process.The Kalman filter method can be used to track the target body in a short time,which is lost,occluded or When other targets are combined,tracking again,the results of the fusion of 2 D lidar and camera information can better obtain the position and boundary of different targets.The research results can provide theoretical reference and basis for the intelligentization of railway safety detection system.

关 键 词：LIDAR 检测 铁路 侵入 

分 类 号：TN985.98[电子电信—信息与通信工程]