An adaptive and efficient selective multiple reference frames motion estimation for H.264 video coding

Yu Ming Lee, Yong Fu Wang, Jia Ren Wang, Yinyi Lin

研究成果: 雜誌貢獻會議論文同行評審

5 引文 斯高帕斯(Scopus)

摘要

In the popular video coding standard H.264/AVC, many advanced techniques are employed. One important technique is the use of multiple reference frames motion estimation. However, the computational load increases with the number of references frames. In this paper, we suggest a selective multiple reference frames motion estimation (SMRFME) architecture which takes use of the information of the 1st reference frame to determine whether it is necessary to search remaining reference frames. In addition, three early termination schemes are applied to the remaining reference frames of the candidate modes. The simulation results demonstrate that the proposed algorithm can achieve up to 77% of time saving compared to the multiple reference frames full search algorithm, while maintaining a high coding performance.

原文???core.languages.en_GB???
頁(從 - 到)509-518
頁數10
期刊Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
5414 LNCS
DOIs
出版狀態已出版 - 2009
事件3rd Pacific Rim Symposium on Image and Video Technology, PSIVT 2009 - Tokyo, Japan
持續時間: 13 1月 200916 1月 2009

指紋

深入研究「An adaptive and efficient selective multiple reference frames motion estimation for H.264 video coding」主題。共同形成了獨特的指紋。

引用此