A New Method for Multiple Target Tracking Based on Fuzzy Possibility

Abstract

In this paper we propose a new method for estimating the time varying number of targets and their states through a sequence of uncertain observations in the presence of clutter and noise, based on fuzzy possibility. The uncertainties of target states and observations at each time step are modeled with fuzzy possibility; furthermore, the birth process of targets and each target survival and detection are modeled with possibility function too. This algorithm estimates the possibility of each target state and new target birth via some fuzzy rules by using fuzzy simulation algorithm. This approach contains a new data association algorithm based on fuzzy possibility for track valued estimate of individual targets. Simulation results show that the proposed algorithm can successfully estimate the state of survival targets, the birth process of targets and the number of targets at each time step in presence of clutter; moreover, it can associate each target with corresponding measurements independently, so this method does not need any additional data association algorithms.