Remote Sensing Images Classification in the city Based on Primary Component Analysis and Fuzzy Neural Network

摘要

This paper deals with the using of Fuzzy Neural Networks (FNNs) to classify the Land-sat 7 image. To raise the speed of classification and solve the redundancy problem, the input neural cells are the next three component of Primary Component analysis, and the NDVI image which used of the planetary albedo to calculate. The precision of evaluation shows that Kappa value of classification is 0.865.