• DocumentCode
    3335023
  • Title

    Multilook technique for dominant scatterer removal in SAR images

  • Author

    Kulpa, K. ; Misiurewicz, J. ; Samczynski, P. ; Smolarczyk, M.

  • Author_Institution
    Inst. of Electron. Syst., Warsaw Univ. of Technol., Warsaw
  • fYear
    2008
  • fDate
    22-24 Sept. 2008
  • Firstpage
    232
  • Lastpage
    235
  • Abstract
    In the paper a multilook method of dominant scatterer removal for SAR image enhancement is presented. Sidelobes produced by very strong scatterers in range and cross-range dimension form a characteristic cross-shaped pattern centred on dominant scatterer. The presence of these sidelobes decreases the overall image quality and can mask weak scatterers close to the strong one. The proposed method is an expanded version of algorithm based on adaptive removal of modelled strong scatterer echoes from raw radar data. The use of multilook technique gives much better results of dasiacleaningpsila the signal. Like in the raw radar signal is reprocessed after removal, and strong scatterers are then added to the final image. The multilook cleaning method has been successfully tested using raw radar data.
  • Keywords
    radar cross-sections; radar imaging; synthetic aperture radar; SAR image enhancement; cross-range dimension form; cross-shaped pattern; dominant scatterer removal; image quality; multilook cleaning method; multilook technique; raw radar signal; scatterer echoes; Cleaning; Image coding; Image enhancement; Image quality; Image resolution; Image segmentation; Radar imaging; Radar scattering; Signal resolution; Testing; SAR image formation; adaptive image enhancement; adaptive removal of dominant scatterer echoes; multilook technique;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Microwaves, Radar and Remote Sensing Symposium, 2008. MRRS 2008
  • Conference_Location
    Kiev
  • Print_ISBN
    978-1-4244-2688-1
  • Electronic_ISBN
    978-1-4244-2689-8
  • Type

    conf

  • DOI
    10.1109/MRRS.2008.4669585
  • Filename
    4669585