Robust and Real Time Object Tracking Using Regional Mutual Information in Surveillance and Reconnaissance Systems

In this paper, a robust object tracking algorithm is proposed based on regional mutual information (RMI) in surveillance and reconnaissance systems. An optimization algorithm is used to object localization in each frame when applying RMI to real time object tracking. In order to improve the performance of optimization, a novel feature point-based background changing detection algorithm is proposed. RMI is suitable as similarity measure for object tracking that reduces sensitivity to noise, partial occlusion and illumination variation. Experimental results demonstrate that our proposed algorithm has high ability to track object when the background changes from un-crowded background to crowded background or vice versa.