多因素观测来源不确定的多目标跟踪算法

Multi-target Tracking Algorithm with Uncertainty of Multi-factor Measurement Origin

杂波、低检测概率和目标间观测相互干扰等因素,使得目标观测的来源难以辨别,提出一种基于改进GMPHD滤波的多目标跟踪算法,通过引入目标标记和权重向量,增广了标准GMPHD迭代中目标信息。基于高斯分量合并策略、目标状态估计策略和高斯分量优化策略,能够有效地改善目标状态估计精度和滤波迭代效率。目标跟踪仿真实验结果表明了所提算法的有效性及鲁棒性。

It is difficult to distinguish the measurement origin of targets due to clutter,low detection probability,and mutual interference among target measurements,etc.therefore,an improved GMPHD filtering based multi-target tracking algorithm is proposed.By introducing the label and weight vector of target,the target information in the standard GMPHD iteration is augmented.Based on the schemes that are Gaussian component merging,target state estimate and Gaussian component optimization,the target state estimation accuracy and filtering iteration efficiency can be improved effectively.The simulation experiment results of target tracking with uncertainty of measurement origin show the effectiveness and robustness of the proposed algorithm.