Linear programming is shown to be a powerful tool for the analysis of mineral reflectance spectra. Algorithms are described, first for stripping the background shape from the diagnostic pattern of narrow absorption features and then for characterizing the main residual features in terms of position, depth, and width. The expected application is automatic recognition of mineral species encountered in mining and exploration contexts, whether for analysis of drill core or for use in smart, field-portable spectrometers.
© 1994 Optical Society of America
Maurice Craig, "Nonconvex hulls for mineral reflectance spectra," Appl. Opt. 33, 849-856 (1994)