Fuzzy Entropy Calculation for SAR Images Classification

Classification methods allow to reduce the dimensionality of a complex data set by grouping them into a set number of classes. In crisp classification, class membership is binary, a sample is a member of a class or not. As well know, this approach is often unsuitable for SAR imagery classification because often the area covered by a region may embrace more than a single class and in this way any information concerning the multi-membership is lost. In fuzzy classification this is not the case while a region can have membership in many different classes to different degrees. In this work we investigate the use of the fuzzy entropy concept for SAR imagery classification. A comparison with standard classification methods highlight the advantages of the proposed approach.