Efficient Key Frame Selection Approach for Object Detection in Wide Area Surveillance Applications

Abstract

This paper presents an efficient preprocessing algorithm for object detection in wide area surveillance video analysis. The proposed key-frame selection method utilizes the pixel intensity differences among subsequent frames to automatically select only the frames that contain the desired contextual information and discard the rest of the insignificant frames. For improving effectiveness and efficiency, a batch updating based on a modular representation strategy is also incorporated. Experimental results show that the proposed key frame selection technique has a significant positive performance impact on wide area surveillance applications such as automatic object detection and recognition in aerial imagery.