Statistical analysis of Synthetic Aperture Radar (SAR) image speckle

Abstract

Synthetic aperture radar (SAR) is an active imaging system that can achieve high resolutions both in range and azimuth. This system has found many real time applications in diverse fields such as geophysics, hydrology, archeology etc. The idea behind SAR is to synthesize the effect of large aperture by moving small aperture radar along the flight path to obtain much finer resolution. While moving along the flight path, the SAR system records reflected waves from the imaged surface at different instants. Coherent processing of these reflected waves from illuminated area of different range and azimuth results in the formation of 2-D SAR images. However, such recorded SAR raw data contains unwanted artifacts which result in granular appearances in SAR image. Those granular parts in SAR image is known as speckle which is multiplicative in nature. Presence of such speckle noise degrade the SAR image quality significantly leading to loss of crucial information. To remove these unwanted signal, statistical properties of speckle needs to be analyzed. So, statistical modeling of the speckle in SAR data plays a key role in the context of SAR image processing and applications. In this research work, we use compound model to separate out the speckle and texture content of recorded SAR raw data. Texture preserving filters such as Lee, Frost, Kuan filter are used to extract texture, and this texture is further used to separate out speckle. Using the estimate of parameters involved in statistical distribution of speckle, the characteristic function method, the approximation of Edgeworth expansion method, and the Method of moments approaches are considered for obtaining the probability density function (pdf) of speckle part of real SAR data obtained from Sandia National laboratory. To validate the usefulness of the above methods for analyzing speckle, we examine the goodness-of-fit between histogram obtained from real SAR data and the pdfs estimated using the above techniques.