Hybrid Approach for Moving Object Detection

摘要

Moving object detection is a major step for video analysis. In this paper, we present a new approach for moving object detection. It is based on motion and edge detection technique. It makes use of the most three recent consecutive frames to detect moving area. Firstly, we compute the infimum gradient and we generate the motion saliency map. Then, we normalize both results to eliminate the parasitic elements. Finally, after applying point-by-point addition between the infimum gradient with the motion saliency map, a morphological closing operation is applied to complete the edge. The experimental results show the effectiveness of our approach for moving object detection with an accuracy rate of 92.49 %.