毫米波雷达点云的高速公路隧道车辆跟踪

Highway Tunnel Vehicle Tracking Based on Point Cloud in Millimeter Wave Radar

针对高速公路隧道车辆跟踪问题,提出一种基于毫米波雷达的低成本解决方案。首先,通过毫米波雷达获取场景目标点云信息,根据每一帧内点云位置、速度以及峰值特点,滤除静态目标与杂波;其次,采用取重心的方法对点云进行匹配,实现感兴趣的车辆目标信息的提取;最后通过相邻帧的时间差预测目标点云下一位置,从而实现不同帧间目标数据匹配跟踪,并与实际场景视频比较。实验结果表明,该算法能够有效解决虚警、漏警问题,提高了单一毫米波雷达传感器测量准确性,实现对车辆目标的跟踪,在跟踪位置上与真实位置非常接近。

Aiming at the problem of highway tunnel scene perception,a low-cost solution based on millimeter wave radar is proposed.Firstly,obtain the point cloud information of the scene target through the millimeter wave radar,filter out the static target and clutter according to the point cloud position,speed and peak characteristics in each frame.Secondly,use the center of gravity method to match the point cloud to achieve interest,the vehicle target information is extracted.Finally,the next position of the target point cloud is predicted through the time difference between adjacent frames,so as to realize the target data matching and tracking between different frames,and compare with the actual scene video.The experiment results show that the algorithm can effectively solve the problem of false alarms and missed alarms,improve the measurement accuracy of a single millimeter-wave radar sensor,and realize the tracking of vehicle targets.The tracking position is very close to the real position.