A Fuzzy Approach Using Euclidean Geometrical Formulation for Classifying SAR Images

Synthetic Aperture Radar (SAR) is a good tool to investigate problems in many geophysical application as classification of ground terrain types and coastal protection. In scientific literature, many analytical and/or numerical techniques have been taken into account to solve the classification problem at hand, especially in all of applications in which it is necessary to classify portion of images with uncertainty and imprecision. In fact, according to the conventional classification approaches, the assignment of a class to each portion of an image could be particularly inadequate for all those portions that span more than a class (for example the coastal areas of the shoreline). This article is devoted to present a fuzzy-geometric approach based on fuzzy subsethood operator to classify SAR images for coastal protection applications. The obtained results were compared, in terms of accuracy, with standard techniques of classification.