Scale-Wavelength Decomposition of Hyperspectral Signals - Use for Mineral Classification & Quantification

An approach for material identification & soil constituent quantification based on a generalized multi-scale derivative analysis of hyperspectral signals is presented. It employs the continuous wavelet transform to project input spectra onto a scale-wavelength space. This allows investigating the spectra at selectable level of detail while normalizing/separating disturbances. Benefits & challenges of this decomposition for mineral classification & quantification will be shown for a mining site.