• DocumentCode
    879473
  • Title

    Multi-frame error concealment for H.264/AVC frames with complexity adaptation

  • Author

    Chen, Chun ; Liu, Yi ; Yang, Zhi ; Bu, Jiajun ; Deng, Xiaoyu

  • Author_Institution
    Coll. of Comput. Sci., Zhejiang Univ., Hangzhou
  • Volume
    54
  • Issue
    3
  • fYear
    2008
  • fDate
    8/1/2008 12:00:00 AM
  • Firstpage
    1422
  • Lastpage
    1429
  • Abstract
    This paper proposes a novel multi-frame error concealment algorithm for the whole-frame losses in video transmission based on H.264/AVC. In order to minimize the error propagation, the proposed algorithm estimates the distortion of both the lost frame and its succeeding frame, and determines the modes for recovering the motion vectors of the lost frame. In addition, based on the analysis of the computational complexity and the distortion of concealed frames, we introduce complexity adaptation into the proposed algorithm to achieve optimal complexity-distortion under different real-time constraints and different decoder computational power. Experimental results demonstrate that the proposed algorithm outperforms conventional algorithms in both objective and subjective quality of the concealed frames.
  • Keywords
    computational complexity; decoding; image motion analysis; video coding; video communication; H.264/AVC frames; computational complexity; decoder computational power; motion vectors; multi-frame error concealment algorithm; multi-frame video coding; optimal complexity-distortion; real-time constraints; video transmission; Algorithm design and analysis; Automatic voltage control; Computational complexity; Decoding; Error correction; Motion estimation; Propagation losses; Streaming media; Video compression; Videoconference; Error concealment, error propagation, H.264/AVC, complexity adaptation;
  • fLanguage
    English
  • Journal_Title
    Consumer Electronics, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0098-3063
  • Type

    jour

  • DOI
    10.1109/TCE.2008.4637636
  • Filename
    4637636