• DocumentCode
    343735
  • Title

    Statistics of heterogeneous terrain at 95 GHz near grazing incidence

  • Author

    De Roo, R.D. ; Ulaby, F.T. ; El-Rouby, A.E.B. ; Nashashibi, A.Y.

  • Author_Institution
    Radiation Lab., Michigan Univ., Ann Arbor, MI, USA
  • Volume
    1
  • fYear
    1999
  • fDate
    11-16 July 1999
  • Firstpage
    52
  • Abstract
    Radar images appear noisy because radar is a coherent measurement process. Even in the absence of thermal noise, radar images exhibit speckle, the salt-and-pepper appearance that shades all of a radar image. While some authors have found it possible to describe sea clutter with such distributions as the K-distribution, we have found that homogeneous terrain clutter is adequately described with the Rayleigh distribution, but that heterogeneous terrain clutter requires the Bayes approach to describe its distribution. This presentation describes briefly the nature of this speckle. The points presented include: speckle from homogeneous clutter is Rayleigh distributed and is determined by a single parameter, the mean return. Speckle from heterogeneous clutter can be considered composed of a distribution of homogeneous clutter with varying mean returns. The Bayes rule can be employed to describe the distribution of speckle for heterogeneous clutter. The value of the mean returns found in nature are tied to the physical properties of the clutter and tend to fall within a certain range of values.
  • Keywords
    Bayes methods; radar clutter; radar imaging; speckle; synthetic aperture radar; 95 GHz; Bayes approach; Rayleigh distribution; SAR image; coherent measurement process; heterogeneous terrain; mean return; near grazing incidence; noise; radar images; speckle; Backscatter; Clutter; Radar imaging; Radar measurements; Radar scattering; Rayleigh channels; Signal resolution; Speckle; Statistics; Voltage;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Antennas and Propagation Society International Symposium, 1999. IEEE
  • Conference_Location
    Orlando, FL, USA
  • Print_ISBN
    0-7803-5639-x
  • Type

    conf

  • DOI
    10.1109/APS.1999.789074
  • Filename
    789074