A video tracking method based on Niche Particle Swarm Algorithm-Particle Filter

In order to improve the stability and robustness in video tracking based on particle filter. We proposed Niche Particle Swarm Algorithm-Particle Filter (NPSA-PF) which applies Niche Particle Swarm Algorithm to the re-sampling stage in particle filter. The ability of Niche Particle Swarm Algorithm which improves the particles´ local search ability and weakens the information sharing between particles, effectively reduces the tracking particles number and improves tracking stability and robustness. We use it in video tracking and the performance is validated to be effective.