Multichannel Classification of SAR-images Using Fuzzy Logic

Abstract

In this paper we show that fuzzy techniques allowing a pixel to belong to more than one class with certain degrees are appropriate classification tools for SAR data. First, we describe a powerful fuzzy classification algorithm. Supervised learning capability allows a high adaptability to user requirements. For many applications satisfying results can be obtained in near real-time. In the second part of the paper we propose a method to fuse classifications obtained from several sources, e.g. different channels of a full polarimetric SAR and/or complementatory classification methods.