A New Frame Based on Particle Filter for Moving Target Tracking

Particle filter is a powerful tool for moving target tracking based on Sequential Monte Carlo methods. But particle filter doesn't take into account the historical prior information on the generation of proposal distribution, and then it cannot approximate posterior density well. A new frame based on particle filter (called FGP-PF) for moving target tracking is proposed in this paper. Firstly, a new prediction model which was based on fuzzy mathematics theory and gray system theory was established, coined FGP (Fuzzy-Grey-Prediction) model. Secondly, the history state sequence is utilized as prior information to predicting and sampled a part of particles for generating proposal distribution in particle filter. Simulation results in a typical motion scenario indicate that the proposed FGP-PF algorithm can exhibits the better overall performance in moving target tracking.