In this paper, a highly flexible model, called the generalized Beta of the first kind (GB1), is proposed to model synthetic aperture radar (SAR) images. Many distributions commonly used to model SAR images are special or limiting cases of the GB1 distribution, including the Gamma, Log-normal, Weibull and the Generalized Gamma distributions. A diagram based on the log-cumulants of the third and fourth orders is proposed to represent and classify these distributions, which could provide an easier-to-visualize characterization of a certain distribution in SAR images. Goodness-of-fit test results using measured SAR data show that the GB1 distribution is a more efficient model to characterize the diversity of scenes in SAR images.
Statistical Modeling of SAR images with the generalized Beta distribution of the first kind
F. GiniPrimo
Membro del Collaboration Group
;M. GrecoMembro del Collaboration Group
;
2017-01-01
Abstract
In this paper, a highly flexible model, called the generalized Beta of the first kind (GB1), is proposed to model synthetic aperture radar (SAR) images. Many distributions commonly used to model SAR images are special or limiting cases of the GB1 distribution, including the Gamma, Log-normal, Weibull and the Generalized Gamma distributions. A diagram based on the log-cumulants of the third and fourth orders is proposed to represent and classify these distributions, which could provide an easier-to-visualize characterization of a certain distribution in SAR images. Goodness-of-fit test results using measured SAR data show that the GB1 distribution is a more efficient model to characterize the diversity of scenes in SAR images.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.