Spatio-temporal context tracking with color attributes

Distinguishing and tracking moving object is one of the most critical tasks in computer vision. A tracker with high speed and precision is required in the real-world object tracking. The correlation filtering tracking approaches have attracted much attention because of its high-speed and robustness. Spatio-Temporal Context (STC) tracker exploits the context information around the target, and improves the tracking performance. For further improving the efficiency and robustness, in the paper, we propose to construct the object appearance using color attributes in STC tracker. Color attributes can be computed efficiently and possess high discriminative power. Experiment results demonstrate that the spatio-temporal context tracker with color attributes can obtain superior performance in real-world tracking.