• DocumentCode
    2468035
  • Title

    Denoising JPEG images by re-application of JPEG

  • Author

    Nosratinia, Aria

  • Author_Institution
    Dept. of Electr. & Comput. Eng., Rice Univ., Houston, TX, USA
  • fYear
    1998
  • fDate
    7-9 Dec 1998
  • Firstpage
    611
  • Lastpage
    615
  • Abstract
    A novel method is proposed for post-processing of JPEG-encoded images, in order to reduce coding artifacts and enhance visual quality. Our method simply re-applies JPEG to the shifted versions of the already-compressed image, and forms an average. This approach, despite its simplicity, offers better performance than other known methods, including those based on nonlinear filtering, POCS, and redundant wavelets
  • Keywords
    data compression; discrete cosine transforms; image coding; image enhancement; noise; transform coding; JPEG image denoising; JPEG re-application; JPEG-encoded images; POCS; already-compressed image; average; coding artifacts reduction; nonlinear filtering; post-processing; redundant DCT expansion; redundant wavelets; shifted versions; visual quality enhancement; Discrete cosine transforms; Filter bank; Image coding; Information filtering; Information filters; Least squares methods; Noise generators; Noise reduction; Quantization; Transform coding;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Multimedia Signal Processing, 1998 IEEE Second Workshop on
  • Conference_Location
    Redondo Beach, CA
  • Print_ISBN
    0-7803-4919-9
  • Type

    conf

  • DOI
    10.1109/MMSP.1998.739048
  • Filename
    739048