State Estimation of Infrared Target using Particle Filter

A filtering algorithm for estimating the position of a moving target using information collected by multiple imaging sensors is presented. Based on camera model and theory of particle filter, state transition model and observation model are set up. Weights threshold is proposed to solve the real-time tracking problem in occlusion issue. The experimental positioning system has been set up to test the algorithm, and the results show that the algorithm is effective in tracking moving target. The estimation error in position of the proposed algorithm is within 0.0082m in tracking stationary target and 0.01m for moving target. State estimation based on PF keeps stable in occlusion condition and is much better than Kalman filter (KF) and extended Kalman filter (EKF) which have large volatility. The experiments results show that the algorithm has good precision and anti-interference performance.