Robust Tracking Based on PSO and On-line AdaBoost

Robust tracking is a challenging problem, due to intrinsic appearance variability of objects caused by in-plane or out-plane rotation and extrinsic factors change such as illumination, occlusion, background clutter and local blur. In this paper, we present a novel framework which uses an on-line AdaBoost algorithm as PSO´s fitness function to search object and attain the efficiency and robustness of the tracking systems. An on-line AdaBoost classifiers based on RGB features is trained and employed. This new tracking framework, which is initialized with the region of the object at the first few frames, can automatically track the object at the remaining frames. We demonstrate our results under some challenging videos, and the results prove our framework is more robust.