复杂交通场景下目标稳定跟踪方法研究

Research on Target Stable Tracking Method in Complex Traffic Scene

摘要:

为克服实际交通场景中环境的复杂性和目标的多样性给雷达目标稳定跟踪带来的问题,采用基于卡尔曼滤波和概率数据关联的雷达多目标跟踪算法,实现点迹与航迹之间的关联。提出一种基于欧氏距离的改进航迹关联方法,增加航向夹角判别的约束,解决了由于多目标相互遮挡或交叉运动带来的航迹不连贯、不稳定问题,实现了航迹与航迹之间的关联,提升了航迹跟踪完整性和稳定性。最后,利用在复杂交通场景下采集的实测交通雷达数据,验证了所提多目标跟踪算法的稳定性和可靠性。

Abstract:

In order to solve the problem that the complexity of environment and the diversity oftargets in actual traffic scene affect stable tracking of radar targets, a radar multi-target trackingmethod based on Kalman filter and probabilistic data association is adopted to correlate point andtrack, An improved track correlation method based on Euclidean distance is proposed, which solvesthe problems of inconsistent and unstable tracking caused by multi-target mutual occlusion or crossmotion by adding constraint on the iudgment of the track ang e. The correlation between tracks isrealized, and the integrity and stability of track tracking is improved, Finally, the stability andreliability of the multi-target tracking algorithm are verified by the measured traffic radar datacollected in complex traffic scenes.