Abstract:

To solve the problem of low realtime tracking caused by the increased particles in a complex environment based on the algorithm of particle filter, the author introduced the background subtraction into the particle filter to propose a new algorithm. After getting the detection result from the image background subtraction, the particle that has been detected to be on foreground pixel is defined as “the important particle”, whose weight grows. Experiments show that the proposed algorithm can achieve better tracking based on fewer particles, and it also improves the realtime tracking.

Key words: artificial intelligence, background subtraction, particle filter, moving target tracking