基于航迹质量的多目标跟踪算法

Research on Track Quality Based Multi-Target Tracking Approach

在多目标跟踪系统中,除了量测与航迹关联问题外,还存在航迹确认和撤销等决策问题。针对这些问题,首先提出了一种航迹质量度量方法,该方法考虑多种不同的检测事件来计算航迹质量;然后在此基础上提出了一种基于航迹质量的多目标跟踪算法,该算法尤其适合存在暂时不可检测目标的场景。根据航迹质量和量测关联历史将航迹分为3个集合:初始航迹、稳定航迹和不可检测航迹。仿真结果表明,基于航迹质量来区分航迹可以提高航迹寿命,同时降低虚假航迹存在时间。

A track quality measure is proposed for assignment-based multi-target trackers. To compute track quality measure for assignment-based data association,one needs to consider different detection events. In addition to the proposed track quality measure,a multi-target tracker based on it is developed,which is particularly suitable in scenarios with temporarily undetectable targets. In this work, tracks are divided into three sets based on their quality and measurement association history： initial tracks,confirmed tracks,and unobservable tracks. The results show that discriminating tracks on the basis of their track quality can lead to longer track life while decreasing the average false track length.