计算机工程与应用 ›› 2014, Vol. 50 ›› Issue (14): 214-218.

• 信号处理 • 上一篇    下一篇

使用动态模型的UMHexagonS算法优化

刘  治,王  玲   

  1. 湖南大学 电气与信息工程学院,长沙 410082
  • 出版日期:2014-07-15 发布日期:2014-08-04

Optimization on UMHexagonS algorithm using dynamic models

LIU Zhi, WANG Ling   

  1. College of Electrical and Information Engineering, Hunan University, Changsha 410082, China
  • Online:2014-07-15 Published:2014-08-04

摘要: H.264取得了很好的编码效率,但是也具有很高的计算复杂度。对H.264中的非对称十字形多层次六边形格点搜索算法(UMHexagonS)进行了优化,分别对提前终止阈值、搜索窗口大小以及搜索模式提出了3种动态模型,提高了算法的自适应性。对六种不同运动程度的视频序列进行了测试,实验结果表明,优化后的算法相对于原来的UMHexagonS算法平均减少了21.67%的编码时间以及47.49%的运动估计时间,同时只有0.02的峰值信噪比下降以及1.69%的比特率增加。

关键词: 非对称十字形多层次六边形格点搜索算法(UMHexagonS), 运动估计, 动态模型, H.264

Abstract: H.264 achieves good coding efficiency but has high computational complexity. This paper proposes a optimization on Unsymmetrical cross Multi Hexagon grid Search(UMHexagonS) algorithm in H.264. Three dynamic models are used in early termination threshold, search range and search mode respectively, which enhance the algorithm’s adaptive ability. Six video sequences with different motion levels are tested in the experiment. Comparing to UMHexagoS algorithm, the proposed algorithm can save 21.67% encoding time and 47.94% motion estimation time on average, with only 0.02 dB degradation in the Peak Signal-to-Noise Ratio(PSNR) and 1.69% increase in the bit rate.

Key words: Unsymmetrical cross Multi Hexagon grid Search algorithm(UMHexagonS), motion estimation, dynamic model, H.264