An alternative derivation of a Bayes tracking filter based on finite mixture models
Liu Wei-feng, Chongzhao Han, Feng Lian · 2009
Abstract – Ba-Tuong-Vo et al proposed a Bayes filter of single target in the random finite set framework [1]. In this paper, we first extend the parameter mixture models (PMM) to state mixture models(s). And further an alterna-tive derivation of a Bayesian tracking filter in clutter is pro-posed for single target. The key of the proposed algorithm is to derive the measurement likelihood function based on finite mixture models. In addition, a closed-form recursion under the linear Gaussian assumption is discussed.