Intelligent modified mean shift tracking using genetic algorithm

Abstract

Object Tracking using mean shift algorithm has gained much attention in recent years due to its simplicity. In this paper, we present a modified mean shift tracking method using genetic algorithm. First, a background elimination method is used to eliminate the effects of the background on the target model. The mean shift procedure is applied only for one iteration to give a good approximate region of the target. In the next step, the genetic algorithm is used as a local search tool to exactly identify the target in a small window around the position obtained from the mean shift algorithm. The simulation results prove that the proposed method outperforms the traditional mean shift algorithm in finding the precise location of the target at the expense of slightly more complexity.