An alternative form of cardinalized PHD filter or I.I.D.-approximation filter
Shozo Mori, Chee-Yee Chong · 2007
In this paper, we derive the updating formula of the cardinalized probability hypothesis density (CPHD) filter recently developed in [1-4], from the non-Poisson multiple-hypothesis tracking (MHT) algorithm developed earlier [23,24]. The particular form of the CPHD updating formula developed in this paper is expressed only with the probability hypothesis density (PHD) or the a posteriori intensity measure density and the a posteriori probability of the number of targets, without using the probability generating function, and is consistent with the updating formula in [5]. Several issues concerning the CPHD updating formula and the sensor modeling are discussed together with a couple of very simple but illustrative examples.