• DocumentCode
    298460
  • Title

    A compound Weibull model for SAR texture analysis

  • Author

    Bucciarelli, T. ; Lombardo, P. ; Oliver, C.J. ; Perrotta, M.

  • Author_Institution
    INFOCOM Dept., Rome Univ., Italy
  • Volume
    1
  • fYear
    34881
  • fDate
    10-14 Jul1995
  • Firstpage
    181
  • Abstract
    Introduces a new model for the statistic of the intensity of synthetic aperture radar (SAR) images. The model, named compound Weibull (CW) distribution, belongs to the well known class of the product models. According to this description the intensity, χ= τ·z, can be interpreted as the product of the variable z named ´speckle´, times the variable τ, referred to as ´texture´. The speckle, representing the effect of the coherent imaging, follows the chi-square distribution with 2L degrees of freedom for an L-look image and mean value 1. The texture, modulating the speckle variance, represents the observed scene and its probability density function (pdf) identifies the specific product model. Its mean value is the mean reflected power and its variance encodes the scene fluctuation. The most well-known product model is the K-distribution obtained with a gamma texture pdf
  • Keywords
    Weibull distribution; geophysical signal processing; geophysical techniques; image texture; radar applications; radar imaging; radar theory; remote sensing by radar; synthetic aperture radar; K-distribution; SAR; chi-square distribution; compound Weibull model; geophysical measurement technique; image texture analysis; intensity; model; probability density function; product model; radar remote sensing; speckle; statistic; synthetic aperture radar; Analytical models; Clutter; Fluctuations; Histograms; Layout; Probability density function; Shape; Speckle; Statistical distributions; Synthetic aperture radar;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Geoscience and Remote Sensing Symposium, 1995. IGARSS '95. 'Quantitative Remote Sensing for Science and Applications', International
  • Conference_Location
    Firenze
  • Print_ISBN
    0-7803-2567-2
  • Type

    conf

  • DOI
    10.1109/IGARSS.1995.519684
  • Filename
    519684