Satellite Image Classification by Fuzzy Neural Network

Abstract

Implementation of Fuzzy Reasoning by structure or Neural Network method is used in this work, weights of Neural Network is Fuzzy Reasoning. The FNN can automatically identify the Fuzzy Rules and tune the membership function by modifying the connection weights of Network using the back-propagation algorithm. Post-classification rules is used to reduce the number of isolated mis-classified pixels occur after the pixel-by-pixel classification principle.