cTracker:一种基于毫米波雷达传感器的室内人员快速检测与追踪系统

cTracker: A fast-indoor people detection and tracking system based on mmWave radar sensor

提出了一种新的基于毫米波雷达传感器的室内人员检测与追踪方法。该方法首先采用一种静态杂波去除算法过滤掉毫米波雷达数据的静态点信息;然后提出了两种有效的聚类算法用于人员目标的聚类与识别。另外,提出了一种递归线性卡尔曼滤波算法(RKF)结合数据关联算法可以同时追踪多个人员。最后,基于提出的算法,实施了一种新的室内人员追踪系统(cTracker),并与现有解决方案—德州仪器(TI)的追踪系统做了对比实验。实验结果显示,该系统的聚类精度从一个人的98%(由错误率计算得来)到5个人的65%不等,在位置1、2、3上的平均位置误差分别为0.2992、0.3271和0.3171 m。相比之下,TI系统的聚类精度从一个人的96%到5个人的45%不等,在位置1、2、3上的平均位置误差分别为0.3283m,0.3116和0.3343 m。综合上述,cTracker系统在跟踪准确率、计算时间和可扩展性方面均优于现有解决方案。

In this paper,a new indoor people detection and tracking system based on millimeter-wave(mmWave)radar sensor is proposed.Firstly,a static clutter removal algorithm is used to remove the static point information in mmWave radar data.Two efficient clustering algorithms are proposed and used for the clustering and identification of the people objects in the scene.Besides,a recursive linear Kalman Filter algorithm is proposed,which,combining with data association algorithm,can be used to track multiple people objects at the same time.Finally,a new fast indoor people detection and tracking system(cTracker)was implemented based on the proposed algorithms,and comparison experiment with the existing solution of the Texas Instruments(TI)system was conducted.The experiment results show that the clustering accuracy of the proposed system is in the range from 98%(calculated from misclassification ratio)for one person to 65%for five people,the average position errors at positions 1,2 and 3 are 0.2992 meters,0.3271 meters and 0.3171 meters,respectively.In comparison,the clustering accuracy of the Texas Instruments system is in the range from 96%for one person to 45%for five people,and the average position errors at positions 1,2 and 3 are 0.3283 meters,0.3116 meters and 0.3343 meters,respectively.The experiment results indicate that the proposed cTracker system is superior to the existing TI system in terms of tracking accuracy,computation time and scalability.