基于改进的概率假设密度多目标跟踪算法

针对多个目标相互紧邻时,概率假设密度滤波器难以正确估计当前目标个数以及目标状态问题,提出一种改进的高斯混合概率假设密度滤波算法。根据每一时刻更新后所有目标的权值构造权值矩阵,通过权值矩阵中目标权值的分布来检测当前目标权值是否存在更新错误。基于新的目标权值再分配策略,对权值矩阵中每个目标可能不正确的权值进行调整,使得每个目标能够获得合理的权值。仿真实验表明,该算法能够准确地估计紧邻目标数目以及状态。

Multi-target Tracking Algorithm Based on Improved Gaussian Mixture Probability Hypothesis Density

For the problem of incorrect estimates of target states and their number in the probability hypothesis density filter when multi-targets move closely each other,an improved Gaussian mixture probability hypothesis density algorithm is proposed. A weight matrix is constructed using the updated weights of all targets at each time step,and the weight update error of targets can be detected based on the weight distribution of the targets in the weight matrix. Based on the novel weight reallocation scheme,the possible incorrect weights of each target in weight matrix are regulated,which makes each target obtain the reasonable weights. Simulation results illustrate that the proposed algorithm can accurately estimate the states of nearby targets and their number.