Robust Visual Tracker via Cascade Classifying

Since the similar background interference and target occlusion, the result of many visual trackers occurred deviation and even lost the target. In this paper, we proposed a robust visual tracker combing with the results of object tracking and target identification to make our algorithm more robust. To do this, we use the cascade classifier to improve the performance of target identification algorithm, then adjust the consequence of object tracking algorithm which based on particle filter. Experiment results show that our method tracks target fast and robustly, and outperform the state-of-the-art tracking methods.