Dimensionality Reduction in Hyperspectral Images

This project includes the dimensionality reduction in hyperspectral images. A fuzzy rough set theory is an approach that deals with the concepts of vagueness as well as indiscernibility and finds the feature subsets preserving the semantics of the given dataset. Therefore, the use of fuzzy rough set method to select the most significant spectral bands from the hyperspectral image is proposed in the given project. The objective of the proposed work is twofold. First, band reduct is performed on hyperspectral image using fuzzy rough set feature selection. Second, hyperspectral image classification is obtained after band reduction. Experiments are carried out with real hyperspectral images acquired by the National Aeronautics and Space Administration Jet Propulsion Laboratory’s Airborne Visible/Infrared Imaging Spectrometer(AVIRIS) and Reflective Optics Spectrographic Imaging System(ROSIS) datasets.