基于车辆数据的k近邻联合概率数据关联算法

A k Nearest Neighbor-Joint Probability Data Association Algorithm Based on Vehicle Data

在交通场景中采用一些预警措施能够有效地减少交通事故发生。例如,对车辆轨迹进行跟踪并预测车辆的驾驶行为,就是一个常用的预警方法。在对车辆进行跟踪的过程中,数据关联是很重要的部分,它可以对车辆的观测点和轨迹进行关联,从而更新车辆的轨迹,完成跟踪过程。在此背景下,提出了一种新的数据关联算法,即k近邻联合概率数据关联算法(k Nearest Neighbor-Joint Probability Data Association,kNN-JPDA)。实验结果表明,该算法能够较好地解决在交通场景下车辆数据的数据关联问题,在精度以及运行效率方面都有所提高。

Some early warning measures can effectively reduce traffic accidents in traffic scenes. For example,tracking vehicle trajectories and predicting vehicle driving behavior is a common early warning method. In the process of vehicle tracking,data association is an important part,which can associate the observations and trajectories of the vehicle,thereby updating the trajectory of the vehicle and completing the tracking process. For this background,a new data association algorithm called k nearest neighbor-joint probability data association algorithm(kNN-JPDA) is proposed. The experimental results show that the algorithm can solve the data association problem of vehicle data in traffic scenes,and it has improved the accuracy and operation efficiency.