A Variational Bayesian algorithm for Extended Target Tracking with Unknown Measurement Noise

The problem of extended target tracking in an unknown non-Gaussian noise field is considered. A Student-t distribution is used to model the measurement noise statistics that combining the sensor error with the extended uncertainty. An improved variational Bayesian algorithm is proposed to estimate the system states and the parameters of the measurement noise. Simulation results show that the proposed method has improved the estimation performance of the extended target tracking system compared with the traditional approaches.