Component pruning based on entropy distribution in Gaussian mixture PHD filter

A component pruning algorithm based on entropy distribution is proposed for the Gaussian mixture implementation of probability hypothesis density filter. Maximum a posterior criterion is adopted for the estimation of mixing parameters. Entropy distribution with negative exponent parameters, which only depends on mixing weights, is adopted as the prior distribution of mixing parameters. The update formulation of mixing weight is derived by Lagrange multiplier. The instability of entropy distribution with negative exponent parameters is applied in driving the components irrelevant with target intensity to extinction during the maximum a posterior iterative procedure. Besides, the problem that an intensity peak is described by several components with similar parameters, can be solved by this instability. It is convenient for the following multitarget state extraction. Simulation results show that the component pruning algorithm based on entropy distribution is superior to the typical algorithm for Gaussian mixture probability hypothesis density filter in multiple target tracking.