Fu, ZhouyuCaelli, TerryLiu, NianjunRobles-Kelly, Antonio2015-12-08August 20-0769525210http://hdl.handle.net/1885/32932Band ratios have many useful applications in hyperspectral image analysis. While optimal ratios have been chosen empirically in previous research, we propose a principled algorithm for the automatic selection of ratios directly from data. First, a robustKeywords: Algorithms; Automation; Classification (of information); Convergence of numerical methods; Image analysis; Large scale systems; Automatic selection; Hyperspectral image classification; Individual ratio features; Kullback-Leibler divergence (KLD); Feature200610.1109/ICPR.2006.3342015-12-08