一种道路条件下车辆跟踪的多目标数据关联方法

A Multi-target Data Association Approach for Vehicle Tracking Under Road Situation

数据关联是多目标跟踪的一个重要组成部分 ,文中介绍了一种基于模糊均值聚类的数据关联方法 ,利用 FCM方法确定权值 ,同时保持 JPDAF方法的基本结构 ,将模糊方法与概率方法有效地结合 .数字仿真表明这种方法有效、简单 。

Data association is an important content in multi-target tracking. Typical algorithms to deal with such problems are the joint probabilities data association filter (JPDAF) proposed by Bar-Shalom and his team. The basis of JPDAF is the calculus of the joint probabilities between the predictions and the measurements. The algorithm assigns weights for reasonable measurements and uses a weighted centroid of those measurements to update the track. In this paper, a new weight assignment method based on fuzzy c-means methodology is proposed, and the general methodology of JPDAF remains unchanged. This leads to a fruitful combination between fuzzy and probability approaches. It is proved that the method is simple and fast by simulation, and adaptive to the multi-target tracking based on automotive radar.