• DocumentCode
    2426742
  • Title

    Fast error resilient H.263 to H.264 video transcoding using RD optimized intra fresh

  • Author

    Liu, Li ; Zhuang, Xinhua

  • Author_Institution
    Comput. Sci. Dept., Univ. of Missouri-Columbia, Columbia, MO
  • fYear
    2008
  • fDate
    7-9 July 2008
  • Firstpage
    370
  • Lastpage
    374
  • Abstract
    A key issue of error resilient transcoding using ldquointra freshrdquo algorithm is how to decide the intra coding rate and adjust corresponding bits for error resilience, so that jointly optimization between video source and error resilience bit can be achieved. To provide practical solution for this issue, we propose to use rate distortion (RD) optimization criterion to decide the application of intra coding mode at MB level. To reduce the implementation complexity of RD optimization intra fresh method, we also present a fast mode decision algorithm that takes advantage of error propagation distortion estimation as assistance to decide if an early termination of existing coding mode choices is necessary. The algorithm is implemented to realize fast error resilient H.263 to H.264 video transcoding. Compared with error resilient transcoding method using random intra update, the proposed algorithm can enhance speed up to 7.2 times while maintaining better picture quality.
  • Keywords
    transcoding; video coding; H.263 video transcoding; H.264 video transcoding; RD optimized intra fresh; error propagation distortion estimation; error resilience; fast mode decision algorithm; intra coding; intra fresh algorithm; rate distortion optimization criterion; Computer errors; Error correction; Lagrangian functions; Optimization methods; Propagation losses; Rate-distortion; Resilience; Streaming media; Transcoding; Video compression;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Audio, Language and Image Processing, 2008. ICALIP 2008. International Conference on
  • Conference_Location
    Shanghai
  • Print_ISBN
    978-1-4244-1723-0
  • Electronic_ISBN
    978-1-4244-1724-7
  • Type

    conf

  • DOI
    10.1109/ICALIP.2008.4590220
  • Filename
    4590220