Spatial-spectral preprocessing for volume-based endmember extraction algorithms using unsupervised clustering

Abstract

Spectral unmixing is an important task in hyperspectral data exploitation. This approach first identifies a collection of spectrally pure constituent spectra, called endmembers, and then expresses the measured spectrum of each mixed pixel as a combination of endmembers weighted by fractions or abundances that indicate the proportion of each endmember present in the pixel. Over the last decade, several algorithms have been developed for automatic extraction of spectral end-members using volume-based concepts. These algorithms use the spectral information contained in the data, and often neglect the spatial information. In this paper, we develop a novel spatial-spectral preprocessing technique for volume-based endmember extraction algorithms intended to exploit spectral information more effectively by adequately incorporating spatial context. Our experimental results, conducted using a real hyperspectral data set collected by NASA's Airborne Visible Infra-Red Imaging Spectrometer (AVIRIS) over the Cuprite Mining district in Nevada, reveal that the proposed approach can successfully integrate the spatial and spectral information contained in the input hyperspectral data.