Fuzzy approach for identifying artistic pigments with Raman spectroscopy

In this work, a fuzzy approach for automatically identifying artistic pigments from their Raman spectra is presented. The uncertainty introduced during the Raman spectrum measurement of pigments is considered in the design of the fuzzy system. The position of the Raman bands in the unknown spectrum can be subject to small displacements due to noise, misalignments in the calibration, etc. Fuzzy logic allows us to work with this uncertainty and to design a system based on the comparison between the Raman band positions in an unknown spectrum recorded from an artwork and the Raman band positions in spectra recorded from reference pigments gathered in databases. The fuzzy system provides the reference pigments whose Raman band positions match those of the unknown pigment analyzed and gives guidance to the decision-making process in the final identification.