Statistical analysis of SAR images†

This is the authors' abstract. We don't add key points for this paper.

Over the last few years various investigations of synthetic aperture radar (SAR) imaging over land targets have been undertaken. Although a great deal of work has been done to develop theoretical models of imaging statistics, most of the actual applications to date have employed SEASAT imagery. Since the 1981 European SAR Convair-580 campaign (SAR-580, CV-580), however, researchers have had a new database of high-resolution SAR imagery with which to work. This paper presents the results of an analysis of the statistical properties of SAR-580 images of various crop types, forest covers and urban scenes. The results of the high-spatial-resolution Convair-580 SAR sensor experiments demonstrate that there are differences in the statistical nature of the different crop, forest and urban scenes studied, and that depending upon the frequency and polarization used during measurement, knowledge concerning the distribution of grey levels in SAR images aids in the discrimination process.