English

应用科学学报 ›› 2002, Vol. 20 ›› Issue (1): 42-46.

• 论文 • 上一篇    下一篇

基于亮度和运动联合分割的位移场估计新算法

汤力, 张兆扬   

  1. 上海大学通信与信息工程学院, 上海 200072
  • 收稿日期:2000-07-13 修回日期:2001-06-11 出版日期:2002-03-31 发布日期:2002-03-31
  • 作者简介:汤力(1976-),男,安徽肥东人,硕士;张兆扬(1938-),男,浙江海宁人,教授,博导.
  • 基金资助:
    国家自然科学基金资助项目(60172020);上海市教委重点项目(S990123)

A New Algorithm for Displacement Field Estimate Based on Joint Luminance and Motion Segmentation

TANG Li, ZHANG Zhao-Yang   

  1. School of Communication and Information Engineering, Shanghai University, Shanghai 200072, China
  • Received:2000-07-13 Revised:2001-06-11 Online:2002-03-31 Published:2002-03-31

PDF

102

被引次数

4

摘要： 运动估计技术是计算机视觉和视频信号处理的关键技术之一,近年来广泛应用于各种视频通信标准中.由于表观运动和实际运动的差异以及孔径问题和遮挡问题的存在,使运动估计难以精确.为此提出一种基于亮度和运动联合分割的位移场估计算法,试图用分割的结果指导运动估计过程,获得了满意的结果.

关键词: 运动估计, 块匹配法, 图象分割, 贝叶斯方法

Abstract: Motion estimate is a key technology in computer vision and video signal processing. It is widely used in international standards for video communication. Due to the difference between real motion and apparent motion and the existence of the aperture and the occlusion problem, it seems difficult to obtain the precise motion field. In this paper, we present a displacement field estimate algorithm based on joint luminance and motion segmentation to obtain a satisfactory result.

Key words: block matching, image segmentation, motion estimate, Bayesian method

中图分类号: