Particle filtering based on compressive sense for target tracking

As for the problems of target blocking and illumination changes in motive target tracking, a particle filtering algorithm based on compressive sense is proposed in this paper. We add the extracted features based on compressive sense of the improved CT algorithm into the framework of particle filtering tracking and judge the credibility of extracted features, as well as the color features of original particle filtering, dealing with the effects of target blocking and illumination changes. The algorithm proposed in this paper is tested in the public database and through experimental results we can find that the algorithm brings about better robust and tracks targets accurately without an increasing calculating complexity, compared with the improved CT algorithm and the particle filtering algorithm.