Spatio-Temporal Context Tracking Algorithm Based on Correlation Filtering

Abstract

The Correlation Filter tracking algorithm is different from the traditional method based on target feature, which has high accuracy and fast tracking speed. In this paper, the Spatio-Temporal Context(STC) target tracking algorithm is combined with Correlation Filtering, the position of the target is predicted by using Correlation Filtering and the prediction result is corrected by using the STC target tracking algorithm. The target tracking algorithm combined with Correlation Filtering and STC can realize effective tracking when the target is occluded or there is an interference target. The experimental results show that the proposed algorithm can not only be applied to the tracking of visual targets in complex backgrounds such as illumination changes, target rotation and background region interference, but also have better robustness.