Multiple Model Student's t Mixture Poisson Multi-Bernoulli Mixture Filter for Multi-Target Tracking with Outliers

Multiple model Poisson multi-Bernoulli mixture (MM-PMBM) filter can achieve stable tracking of multiple maneuvering targets. However, the MM-PMBM filter models the process and measurement noise as Gaussian distribution and does not consider the scenario where some outliers existing in the dynamic system or the measurement system. To solve this problem, a multiple model Student's t mixture Poisson multi-Bernoulli mixture (MM-St-PMBM) filter is proposed in this paper. Firstly, in the proposed filter, the process and measurement noises are modeled as Student's t-distributions. Secondly, by making full use of the characteristics of the Student's t-distribution and the MM-PMBM filter, the proposed filter approximates the multi-target intensity as Student's t mixture components to be propagated in the estimation process. Finally, the tracking effectiveness of the MM-St-PMBM filter in the complex multi-maneuvering target tracking with noise outliers scenario is shown by simulation experiments.