• DocumentCode
    1880707
  • Title

    Enhanced adaptive intra mode bit skip for H.264/AVC

  • Author

    Po, Lai-Man ; Wang, Liping ; Feng, Litong ; Ng, Ka-Ho ; Wong, Ka-Man ; Cheung, Kwok-Wai

  • Author_Institution
    Dept. of Electron. Eng., City Univ. of Hong Kong, Hong Kong, China
  • fYear
    2012
  • fDate
    12-15 Aug. 2012
  • Firstpage
    98
  • Lastpage
    103
  • Abstract
    Adaptive intra mode bit skip (AIMBS) can improve the coding efficiency of intra coding in H.264/AVC. However, it has two shortcomings: one is that DC mode is not effectively handled in AIMBS as DC mode is duplicated in both Single-Prediction and Multiple-Prediction processing; the other is that AIMBS does not well match the most probable mode (MPM) estimation scheme of H.264/AVC. To tackle these two shortcomings, an enhanced AIMBS (EAIMBS) scheme with larger bitrate reduction and better QP robustness is presented. In addition, EAIMBS almost maintains as low computational complexity as AIMBS. The proposed technique introduces distance-based weighted prediction (DWP) to replace DC mode in Multiple-Prediction and it calculates the MPM by simplified L-shaped most probable mode estimation (SLMPME) method. Experimental results show that EAIMBS gives in average 3.97% bit-rate reduction and 51% computation reduction at QP=32 compared with H.264/AVC.
  • Keywords
    computational complexity; video coding; AIMBS; DWP; H.264/AVC; SLMPME; bitrate reduction; coding efficiency; computation reduction; computational complexity; distance-based weighted prediction; enhanced adaptive intra mode bit skip; intra coding; most probable mode estimation scheme; multiple prediction processing; simplified L-shaped most probable mode estimation; single prediction processing; Accuracy; Bit rate; Cities and towns; Encoding; Estimation; PSNR; Video coding; AIMBS; EAIMBS; H.264; Intra coding;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Signal Processing, Communication and Computing (ICSPCC), 2012 IEEE International Conference on
  • Conference_Location
    Hong Kong
  • Print_ISBN
    978-1-4673-2192-1
  • Type

    conf

  • DOI
    10.1109/ICSPCC.2012.6335588
  • Filename
    6335588