• DocumentCode
    3398780
  • Title

    Collaborative filtering technique for SAR image speckle noise suppression

  • Author

    Gromek, Artur ; Castaldo, Luigi

  • Author_Institution
    Inst. of Electron. Syst., Warsaw Univ. of Technol., Warsaw, Poland
  • fYear
    2013
  • fDate
    5-7 June 2013
  • Firstpage
    1
  • Lastpage
    4
  • Abstract
    Collaborative filtering is a technique used for image restoration among noise and distortions. Based on the fact that an image content has a locally sparse representation in transform domain. Technique employs a non-local modeling of image by collecting similar 2D image patches into 3D groups. The so-called collaborative filtering applied on such a 3D groups is realized by transform domain shrinkage and Wiener filtration. In this work, we implement collaborative filtering algorithm and test it on Synthetic Aperture Radar (SAR) imagery. The experimental results will be presented.
  • Keywords
    Wiener filters; collaborative filtering; image representation; image restoration; radar imaging; synthetic aperture radar; 2D image patches; SAR image speckle noise suppression; Wiener filtration; collaborative filtering technique; image content; image restoration; sparse representation; synthetic aperture radar; transform domain; transform domain shrinkage; Collaboration; Filtering; Noise; Speckle; Synthetic aperture radar; Three-dimensional displays; Wiener filters; SAR; despeckling; non-local filtering;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Signal Processing Symposium (SPS), 2013
  • Conference_Location
    Serock
  • Type

    conf

  • DOI
    10.1109/SPS.2013.6623570
  • Filename
    6623570