The Precise Recognition of Moving Object in Complex Background

Abstract

Moving object detection is currently one of the most active research topics in the domain of computer vision and video processing. In this paper, a simply and fast quadric binarization method is proposed to remove environmental noise points effectively. Moreover, based on the stability of drawing the contour from fixed scene, the background subtraction method is combined with the time-stepping method, an effective method which could be used to detect the movement areas. Through the average weighted of two means can realize to enhance the accuracy in distinguishing target. Experiment results have shown that this method gives stable performances and good robustness.