Combining Color Attributes for Scale Adaptive Correlation Tracking

Correlation filter-based trackers achieve very good performance in visual tracking, but the traditional correlation tracking methods failed in mining the color information of the image sequence. To solve this problem, we propose a novel and robust scale adaptive tracker combined with color attributes in correlation filter framework, which extracts not only gray but also color information as the feature maps to compute the maximum response location via multi-channel correlation filters. Furthermore, we employ a scale pyramid to estimate the target scale. Experiments are performed on several benchmark challenging color sequences. And the results show that our approach outperforms state-of-the-art tracking methods while operating in real-time.