A Novel Particle Filter Based Background Subtraction Method
Enwei Zhang, Feng Chen, Weidong Zhang · 2006
Background subtraction is a typical method to segment foreground from background in motion analysis. Most of background subtraction methods base on static backgrounds. We propose a particle filter based method to estimate dynamic backgrounds, such as water waves, foliage, smoke, etc. We consider the dynamic scenes vary under some rules and use particle filter to track background. Experiments show our results are promising, compared with other background subtraction methods