计算机工程与应用 ›› 2010, Vol. 46 ›› Issue (33): 139-141.DOI: 10.3778/j.issn.1002-8331.2010.33.039

• 数据库、信号与信息处理 • 上一篇    下一篇

用于运动估计的基于梯度下降搜索扩展算法

阴法明1,赵晓铃2   

  1. 1.南京信息职业技术学院,南京 210013
    2.大庆石油管理局通信公司,黑龙江 大庆 163453
  • 收稿日期:2009-10-23 修回日期:2009-12-23 出版日期:2010-11-21 发布日期:2010-11-21
  • 通讯作者: 阴法明

Extended block-based gradient descent search algorithm for motion estimation

YIN Fa-ming1,ZHAO Xiao-ling2   

  1. 1.Nanjing Information Occupation Technical College,Nanjing 210013,China
    2.Communication Company of Daqing Petroleum Administration Bureau,Daqing,Heilongjiang 163453,China
  • Received:2009-10-23 Revised:2009-12-23 Online:2010-11-21 Published:2010-11-21
  • Contact: YIN Fa-ming

摘要: 在快速搜索算法中,基于梯度下降搜索算法对搜索速度的改善非常明显,但搜索质量下降较大,尤其是对于帧间图像运动较大的序列。对该算法进行针对性扩展,提出基于梯度下降搜索扩展算法。与全搜索、新三步搜索、四步搜索、基于块的梯度下降搜索、钻石搜索算法相比较,实验结果表明该算法在搜索速度和质量上表现良好。

关键词: 视频编码, 运动估计, 块匹配算法, 梯度

Abstract: Among the fast search algorithms,Block-Based Gradient Descent Search(BBGDS) algorithm improves the search speed obviously.However,the decline of search quality is comparatively large,especially for the sequences which have a large amount of movements.This paper extends this algorithm pertinently and proposes an extended BBGDS(EBBGDS) algorithm.The EBBGDS is compared with Full Search(FS),New Three-Step Search(NTSS),Four-Step Search(FSS),Block-based Gradient Descent Search(BBGDS),and Diamond Search(DS).Experimental results show that the proposed algorithm provides competitive performance in search speed and quality.

Key words: video coding, motion estimation, block matching algorithm, gradient

中图分类号: