Using PSO Particle Filter to Recovery Image
Haitao Yao, Fuxi Zhu · 2011
Particle filter has the advantage of processing nonlinear and non-Gaussian problems in complex motion estimation. This paper describes a new image recovery algorithm based on PSO (particle swarm optimization) Particle Filter. The PSO is introduced into particle filter to improve convergence speed and the capacity of particles moving to high likelihood area, thus high estimation accuracy of particles can be achieved. The experimentation for mixed noise images indicates the validity and superiority of this algorithm.