Due to the fact that the unmatching of extended targets’ measurements noise is also assumed as Gaussian distribution and conventional algorithm cannot estimate target’s extent under the circumstance of unknown measurement noise covariance, a new multiple extended target tracking algorithm based on variational bayesian random hypersurface model (VB-RHM) is proposed, which is embedded into CPHD filter frame. Measurement noise is modeled by glint noise with t-distribution, its parameters are assumed to have a Gamma prior distribution so that the predicted and updated PHDs can have mixture of Gaussians representations. A variational bayesian procedure is applied to iteratively estimate parameters of the mixture distributions through random hypersurface model CPHD prediction and update steps. The simulation results show that the proposed algorithm VB-RHM-CPHD can track multiple extended targets’ kinematic state and object extension under the condition of unknown numbers and measurement noise covariance adaptively. In addition, it has an improved precision compared with conventional RHM-GGM-CPHD algorithm.
