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

Guifen Tian, Tianruo Zhang, Takeshi Ikenaga, Satoshi Goto

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

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.

Original languageEnglish
Title of host publicationAdvances in Multimedia Modeling - 15th International Multimedia Modeling Conference, MMM 2009, Proceedings
Pages85-95
Number of pages11
DOIs
Publication statusPublished - 2009 Feb 5
Event15th International Multimedia Modeling Conference, MMM 2009 - Sophia-Antipolis, France
Duration: 2009 Jan 72009 Jan 9

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5371 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

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

Keywords

  • Block Size Decision
  • Entropy Theory
  • H.264/AVC
  • Intra Prediction
  • Mode Decision

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'A fast hybrid decision algorithm for H.264/AVC intra prediction based on entropy theory'. Together they form a unique fingerprint.

  • Cite this

    Tian, G., Zhang, T., Ikenaga, T., & Goto, S. (2009). A fast hybrid decision algorithm for H.264/AVC intra prediction based on entropy theory. In Advances in Multimedia Modeling - 15th International Multimedia Modeling Conference, MMM 2009, Proceedings (pp. 85-95). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5371 LNCS). https://doi.org/10.1007/978-3-540-92892-8_10