A fast hybrid decision algorithm for H.264/AVC intra prediction based on entropy theory

Guifen Tian, Tianruo Zhang, Takeshi Ikenaga, Satoshi Goto

研究成果: Conference contribution

3 被引用数 (Scopus)

抄録

Rate Distortion Optimization based spatial intra coding is a new feature of H.264/AVC standard. It efficiently improves the video coding performance by brutally utilizing variable block sizes and multiple prediction modes. Thus, extremely high computation complexity is required. This paper proposes a hybrid decision algorithm which can reduce unimportant block sizes and prediction modes. Entropy feature of each MB is extracted to decide the optimal block size. Then Prewitt operators based edge direction extraction module is designed to select only promising prediction modes. In addition, above two decision modules can function together to form a fast hybrid decision algorithm for intra prediction. Sufficient experiments prove that this hybrid algorithm is able to achieve 63% time reduction with PSNR loss limited to 0.07dB, bit rate increase at most 2.02%. It is expected to be utilized as a favorable accelerator for a real-time H.264 encoder for HDTV.

本文言語English
ホスト出版物のタイトルAdvances in Multimedia Modeling - 15th International Multimedia Modeling Conference, MMM 2009, Proceedings
ページ85-95
ページ数11
DOI
出版ステータスPublished - 2009 2 5
イベント15th International Multimedia Modeling Conference, MMM 2009 - Sophia-Antipolis, France
継続期間: 2009 1 72009 1 9

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
5371 LNCS
ISSN(印刷版)0302-9743
ISSN(電子版)1611-3349

Conference

Conference15th International Multimedia Modeling Conference, MMM 2009
CountryFrance
CitySophia-Antipolis
Period09/1/709/1/9

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

フィンガープリント 「A fast hybrid decision algorithm for H.264/AVC intra prediction based on entropy theory」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル