A Visual Object Tracking Method Based on Improved Bhattacharyya Coefficient and Model Update Strategy
Huang An-Qi, Hou Zhi-Qiang, Wangsheng Yu, Xiang Liu · 2014
In the traditional Mean Shift tracking algorithm, The Bhattacharyya coefficient is an efficient method in image statistical feature matching. But for the influence of background feature, the optimal location obtained by Bhattacharyya coefficient may not be the exact target location. Thus, there will be drifted or even wrong location in tracking. This paper proposes an improved Bhattacharyya coefficient based on visual saliency. The new coefficient effectively reduces the influence of background feature, and emphasizes the importance of target feature, which obviously improves the target matching accuracy compared to the original coefficient. In order to get an effective model update strategy, the paper synthetically analyzes the similarity of target model and background model, and estimates the reason of the disturbance. The experiment result shows the proposed method can well restrain background distraction, meanwhile, it can effectively update the model and overcome the problem of model drift, and the tracking algorithm is effective and robust.