Study On The Extraction Of Moving Objects Based On The Calculation Model Of Contour Grouping

Contour grouping calculation model use the edge fragment as grouping element. The main task is to extract meaningful object contour and it’s a very important calculation model of perceptual organization. The Significant moving object extraction algorithm by means of contour grouping calculation model can better uses the principles of perception organization and provide more accurate inputs for the follow-up work such as target recognition and scene understanding and so on. Contour grouping calculation model has important theoretical value and practical significance.This paper is based on the research of the contour grouping calculation model and explores the temporal and spatial characteristics of the grouping element and the salience of the closed contour. Based on this, we have implemented the extraction of moving objects in image sequences Its main works are presented as follows:(1) Through the research on the motion characteristics of edge segments in the image sequence, in this paper, an optimization algorithm of grouping elements based on global motion contrast is presented. This algorithm uses the motion information of the grouping element’s neighborhood to distinguish between target and non-target edge fragment. The experiment of grouping element extraction in the dataset indicate that compared with the existing algorithm, this algorithm can obtain better results.(2) The study focuses on regional motion saliency in the image sequence. From the definition of the significant contour grouping element, we studied the influence of the regional motion saliency for contour grouping. This paper presents a contour grouping algorithm based on the significant motion. Using the regional motion saliency as the Constraint of edge fragment connection to reduce the Influence of non-salient regions for contour grouping. Experiments indicate that this algorithm can improve the quality of moving object contour extraction.(3) The study focuses on the regional motion feature. In this paper, a contour grouping algorithm based on regional motion affinity. Using the regional motion affinity as the constraint of grouping cost function. Through the optimization algorithm to extract the closed contour. Experiments indicate that this algorithm can improve the quality of moving object contour extraction.