Computer Engineering and Applications ›› 2008, Vol. 44 ›› Issue (17): 185-187.

• 图形、图像、模式识别 • Previous Articles     Next Articles

Fast subpel motion estimation algorithm based on group error

XIANG Dong1,LUO Zheng-hua2   

  1. 1.School of Computer Science and Technology,Wuhan University of Technology,Wuhan 430063,China
    2.Information Academy,Zhongnan University of Economics and Law,Wuhan 430073,China
  • Received:2007-09-02 Revised:2007-11-28 Online:2008-06-11 Published:2008-06-11
  • Contact: XIANG Dong

基于分组误差的快速分数像素运动估计算法

向 东1,骆正华2   

  1. 1.武汉理工大学 计算机科学与技术学院,武汉 430063
    2.中南财经政法大学 信息学院,武汉 430073
  • 通讯作者: 向 东

Abstract: H.264 is the latest international video coding standard,it supports high accurate subpel motion estimation.When the fast integer motion estimation is used,the computational load of subpel motion estimation may be the bottleneck of motion estimation.According to the monotonically increasing match criteria around the location of the optimal motion vector a fast fractional motion estimation algorithm based on the group error is presented.Simulation results show that it can save the search points of subpel motion estimation efficiently with very low rate distortion loss.

摘要: H.264是最新一代的视频编码标准,它支持高精度分数像素运动估计。当在整数运动估计中采用快速算法后,分数像素的运动估计可能成为制约快速运动估计算法性能的瓶颈。根据匹配准则的目标函数在最优运动矢量的附近都是单调递增这一假设,提出一种基于分组误差的快速分数像素运动估计算法。实验结果表明该算法不但能够有效降低分数像素运动估计的计算量,而且还能保持较小的率失真性能损失。