ABSTRACT
In the multisensor tracking system, the objective of multitarget tracking is to estimate the states of targets at each time step from the sequence of noisy and cluttered observation sets obtained by each sensor. In 2003 Mahler employed the random set framework to propose a probability hypothesis density (PHD) filter. This method can avoid the data association between observations and objects. Then the sequential Monte Carlo (SMC) method is used to implement the PHD filter. This chapter discusses the problem of analytic implementation of the exact MS-PHD filter. It achieves the analytic implementation of exact MS-PHD formulas by using Gaussian mixture under the linear Gaussian system assumptions. The chapter proposes a heuristic partition method to reduce the computational complexity of the analytic implementation of the exact MS-PHD corrector. It then adopts the random finite set (RFS) framework of Vo's to model the multisensor multiobject tracking problem.
