• DocumentCode
    147201
  • Title

    A low power architecture for H.264 encoder in Intra Prediction Mode

  • Author

    Senthilkumar, G. ; Hariprasath, S.

  • Author_Institution
    Dept. of Electron. & Commun. Eng., Saranathan Coll. of Eng., Trichy, India
  • fYear
    2014
  • fDate
    3-5 April 2014
  • Firstpage
    1486
  • Lastpage
    1490
  • Abstract
    H.264 requires large number of computational elements due to their computational complexity. By reducing this computational complexity the low power architecture is proposed in this paper. In this work the computational reduction using Pixel similarity computation and the Minimum SAD computation is proposed. The output of pixel similarity computation reduces the number of comparisons between current block pixels and the neighboring block pixels in Intra Prediction Mode. As a result the computational complexity in terms of additions and multiplications are reduced in the design of encoder and consequently the power consumption is also reduced. In this work the 4×4 luma block is taken into account for the pixel comparison and residual prediction. The power consumption for the architecture is reduced by 22.77%. The architecture is implemented in Verilog. The RTL for the Verilog code is verified in Virtex5 XC5VLX50T FPGA.
  • Keywords
    data compression; prediction theory; video codecs; video coding; H.264 encoder; Pixel similarity computation; Verilog code; Virtex5 XC5VLX50T FPGA; computational complexity; computational elements; computational reduction; intra prediction mode; low power architecture; luma block; minimum SAD computation; Computational complexity; Computer architecture; Encoding; Equations; Power demand; Prediction algorithms; Standards; H.264; Intra Prediction; Low power; Minimum SAD(Sum of Absolute Difference); Pixel Similarity;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Communications and Signal Processing (ICCSP), 2014 International Conference on
  • Conference_Location
    Melmaruvathur
  • Print_ISBN
    978-1-4799-3357-0
  • Type

    conf

  • DOI
    10.1109/ICCSP.2014.6950096
  • Filename
    6950096