Multiple Feature Fused for Visual Tracking via Correlation Filters

The general tracking algorithm is vulnerable to noise because of using a single feature, makes the performance and robustness of the those algorithms greatly limited. In this paper, in order to achieve the robust and pretty performance, we propose a novel multiple feature fused model in correlation filter framework for visual tracking. The adoption of complementarity between different features can effectively eliminate the effects of noise and maintain their advantages of different features. While the correlation filter framework can provide a fast training and locate mechanism. In addition, we give a simple but effective scale detection method, which can appropriately handle the scale variation in the tracking sequences. We evaluate our tracker on OTB2013 benchmark, which include 51 video sequences. On this dataset, our results show that the proposed approach achieves a promising performance.