Classification of multispectral satallite images by using adaptive neuro-fuzzy classifier with linguistic hedges

In this study, vegetation species were classified by using multispectral satellite images. A full wavelet transform is used to decompose the images into sub-images and the energy in each sub-images is assigned as feature for classification. These features were eliminated and classified by using neuro-fuzzy classifier with linguistic hedges. A classification accuracy of 93.75% was achieved by using the selected five features among 252 extracted features.