• DocumentCode
    1911800
  • Title

    Grid Based Method for Computing DCT Coefficients in Arbitrary Rectangle Region

  • Author

    Lin Xiao ; Xiao Guoqiang

  • Author_Institution
    Comput. & Inf. Sci. Dept., Southwest Univ., Chongqing, China
  • fYear
    2012
  • fDate
    14-16 Dec. 2012
  • Firstpage
    456
  • Lastpage
    459
  • Abstract
    At present, almost all digital images are transferred to the compressed format for storing and communication reason, in which the discrete cosine transform (DCT) based compression method remains an effective and important way for image compression because of the effort from JPEG. It is considerable to introduce image processing techniques such as feature extraction, pattern recognition and image indexing to the compressed image domain. In these image processing techniques, they all involve operations on arbitrary regions in images. So it is critical to find an effective way to calculate the DCT coefficients in arbitrary regions. To this end, we propose a grid based approach to find the most suitable relationship between DCT coefficients in 8*8 macro-blocks and arbitrary rectangle region. Experiment result reveals that the proposed method has a lower computational complexity than the butterfly algorithm (BA) when computing the DCT coefficients in arbitrary rectangle region.
  • Keywords
    computational complexity; data compression; discrete cosine transforms; grid computing; image coding; DCT coefficients; JPEG; arbitrary rectangle region; butterfly algorithm; compressed format; compressed image domain; computational complexity; digital image; discrete cosine transform based compression; feature extraction; grid based method; image compression; image indexing; image processing; pattern recognition; 2D-DCT; arbitrary-recetangle; computation complexity; grid;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Information Science and Engineering (ISISE), 2012 International Symposium on
  • Conference_Location
    Shanghai
  • ISSN
    2160-1283
  • Print_ISBN
    978-1-4673-5680-0
  • Type

    conf

  • DOI
    10.1109/ISISE.2012.109
  • Filename
    6495386