Texture and speckle in high resolution synthetic aperture radar clutter

Abstract

A theoretical model of high-resolution synthetic aperture radar (SAR) clutter that is able to predict the statistical effects of spatial averaging upon homogeneous clutter that is speckled and spatially correlated is discussed. The predictions of this model, when the texture correlation was assumed to be Gaussian-shaped, were found to be in excellent agreement with experimental results using grass clutter seen through a high-resolution radar. The predictions of this model, when the texture correlation was assumed to be exponential-shaped with a slant-range cosine factor, were found to be in excellent agreement with experimental results using tree clutter seen through a high-resolution radar. The use of an exponential shape, rather than a Gaussian shape, was necessary in order to achieve the more gradual decay in spatial correlations observed with tree clutter. The use of a slant-range cosine factor was necessary in order to achieve the negative correlations that were observed with tree clutter in the slant-range, but not the cross-range, dimension. >