New developments in target detection in hyperspectral imagery using spectral metrics and spectra extraction

An important problem in processing multispectral / hyperspectral imagery consists in the design of methods for the detection of targets. In this paper we investigate a class of detection filters based on band and spectral selection. Band selection (band screening) refers to searching for the spectral bands that would yield the largest separation between target and background. Spectral selection (spectral screening) is a technique used in reducing the multispectral / hyperspectral data to a representative subset of spectra. The subset is formed such that any two spectra in it are dissimilar and, for any spectrum in the original image cube, there is a similar spectrum in the subset. Spectral screening is performed in a sequential manner, at each step, the subset being increased with a spectrum dissimilar from all the spectra already selected. We modified the spectral screening algorithm such that at the selection step the spectrum with the largest distance from the set is selected. While not introducing additional computational complexity, the Maximum Spectral Screening (MSS) algorithm ensures that the overlap among the representatives is minimized. The detection filters were obtained as the classification projector matrices based on the spectral subset. The developed algorithms were tested on HYDICE hyperspectral data using the spectral angle and the spectral information divergence. The results indicate that regular MSS outperforms band selection followed by MSS, both classes outperforming regular spectral screening.