基于时空关联—网格聚类的多扩展目标跟踪算法
|
摘要
针对多扩展目标的跟踪问题,提出了基于时空关联—网格聚类的多扩展目标跟踪算法。根据时空相关性,将当前量测分为存活目标量测和新生目标量测。对存活目标,利用模糊C均值(FCM)算法进行量测划分,由高斯混合—概率假设密度(GM-PHD)滤波器得到存活目标轨迹。对新生目标,用网格聚类完成量测集划分,由扩展目标—高斯混合—概率假设密度(ET-GM-PHD)滤波器得到新生目标的轨迹。仿真结果表明:所提算法能够对多扩展目标进行准确跟踪,特别是在目标交叉时刻,估计目标数目更准确,算法实时性更好。
Aiming at the problem of multiple extended target tracking,a multiple extended target tracking algorithm based on spatiotemporal correlation and grid clustering is proposed. The current measurements can be divided into measurements of survival targets and newborn targets according to spatiotemporal correlation. For survival targets,the measurements are partitioned by fuzzy C-means( FCM) algorithm,the trajectories of survival targets are received by Gaussian mixture probability hypothesis density( GM-PHD) filter. For newborn targets,the measurement sets are partitioned by grid clustering,the trajectories of newborn targets are received by ET-GMPHD filter. Simulation results show the proposed algorithm can track multiple extended target accurately,especially at the moment of targets crossing estimated number of targets,is more accurate and realtime performance of algorithm is better.
Aiming at the problem of multiple extended target tracking,a multiple extended target tracking algorithm based on spatiotemporal correlation and grid clustering is proposed. The current measurements can be divided into measurements of survival targets and newborn targets according to spatiotemporal correlation. For survival targets,the measurements are partitioned by fuzzy C-means( FCM) algorithm,the trajectories of survival targets are received by Gaussian mixture probability hypothesis density( GM-PHD) filter. For newborn targets,the measurement sets are partitioned by grid clustering,the trajectories of newborn targets are received by ET-GMPHD filter. Simulation results show the proposed algorithm can track multiple extended target accurately,especially at the moment of targets crossing estimated number of targets,is more accurate and realtime performance of algorithm is better.
引文
[1]刘妹琴,兰剑.目标跟踪前沿理论与应用[M]].北京:科学出版社,2015:202-203.
[2]刘丽娟,刘国栋.基于ET-GM-PHD的多传感器多目标跟踪算法[J].传感器与微系统,2013,32(10):126-133.
[3]多传感器多目标跟踪的粒子PHD滤波算法[J].传感器与微系统,2010,29(4):18-25.
[4] Mahler R. PHD filters for nonstandard targets,I:Extended targets[C]∥Proc of the 12th International Conference on Information Fusion,2009:915-921.
[5] Granstr9m K,Lundquist C,Orguner U. A Gaussian mixture PHD filter for extended target tracking[C]∥2010 the 13th Conference on Information Fusion,IEEE:2010:1-8.
[6] Granstrom K,Lundquist C,Orguner U. Extended target tracking using a Gaussian mixture PHD filter[J]. IEEE Trans on Aerospace and Electronic System,2012,48(4):3268-3286.
[7]韩玉兰,任重义,韩崇昭.多扩展目标滤波器的量测集划分算法[J].压电与声光,2015,37(4):603-608.
[8]蒲磊,冯新喜,孔云波,等.基于SNN-AP聚类的扩展目标量测集划分方法[J].计算机应用研究:网络优先版,2017,34(5):1349-1352.
[9]章涛,吴仁彪.近邻传播观测聚类的多扩展目标跟踪算法[J].控制与决策,2016,31(4):764-768.
[10]曹倬,冯新喜,蒲磊,等.基于高斯混合概率假设密度滤波器的扩展目标跟踪算法[J].系统工程与电子技术,2017,39(3):494-499.
[11]何熊熊,管俊轶,叶宣佐,等.一种基于密度和网格簇心可确定聚类算法[J].控制与决策,2017,32(5):913-919.
[12] Schuhmacher D,B T Vo,B N Vo. A consistent/metric for performance evaluation of multi-object filter[J]. IEEE Transactions on Signal Processing,2008,56(8):3447-3457.
[2]刘丽娟,刘国栋.基于ET-GM-PHD的多传感器多目标跟踪算法[J].传感器与微系统,2013,32(10):126-133.
[3]多传感器多目标跟踪的粒子PHD滤波算法[J].传感器与微系统,2010,29(4):18-25.
[4] Mahler R. PHD filters for nonstandard targets,I:Extended targets[C]∥Proc of the 12th International Conference on Information Fusion,2009:915-921.
[5] Granstr9m K,Lundquist C,Orguner U. A Gaussian mixture PHD filter for extended target tracking[C]∥2010 the 13th Conference on Information Fusion,IEEE:2010:1-8.
[6] Granstrom K,Lundquist C,Orguner U. Extended target tracking using a Gaussian mixture PHD filter[J]. IEEE Trans on Aerospace and Electronic System,2012,48(4):3268-3286.
[7]韩玉兰,任重义,韩崇昭.多扩展目标滤波器的量测集划分算法[J].压电与声光,2015,37(4):603-608.
[8]蒲磊,冯新喜,孔云波,等.基于SNN-AP聚类的扩展目标量测集划分方法[J].计算机应用研究:网络优先版,2017,34(5):1349-1352.
[9]章涛,吴仁彪.近邻传播观测聚类的多扩展目标跟踪算法[J].控制与决策,2016,31(4):764-768.
[10]曹倬,冯新喜,蒲磊,等.基于高斯混合概率假设密度滤波器的扩展目标跟踪算法[J].系统工程与电子技术,2017,39(3):494-499.
[11]何熊熊,管俊轶,叶宣佐,等.一种基于密度和网格簇心可确定聚类算法[J].控制与决策,2017,32(5):913-919.
[12] Schuhmacher D,B T Vo,B N Vo. A consistent/metric for performance evaluation of multi-object filter[J]. IEEE Transactions on Signal Processing,2008,56(8):3447-3457.