Using local saliency for object tracking with particle filters

Particle filter shows great success in non-rigid object tracking. But tracking errors would be inevitable when the background changes greatly, or when the color distribution of background is similar to target´s. To eliminate these errors, an improved method based on visual attention is proposed in this paper. In our algorithm, the extraction method based on visual attention is applied to extract salient region from target area and particle areas. Instead of saliency map of the whole image, we just compute saliency map of each particle to extract salient region. Experimental results show that the proposed method is more robust against background interference and illumination changes. Therefore, the target tracking is more stable.