Fuzzy logic for identifying pigments studied by Raman spectroscopy

Area of Science:

• Analytical Chemistry
• Art Conservation Science
• Computational Materials Science

Background:

• Raman spectroscopy is crucial for analyzing pigments in cultural heritage.
• Accurate pigment identification is vital for art conservation and authentication.
• Automated analysis methods are needed to process complex spectral data.

Purpose of the Study:

• To develop and validate an automated method for interpreting Raman spectra of pigments using fuzzy logic.
• To create a reference library of pigment spectra for automated analysis.
• To assess the algorithm's performance on historical pigments and mixtures.

Main Methods:

• Utilized fuzzy logic and linguistic variables for spectral interpretation.
• Developed a methodology for extracting featured bands and constructing a pigment reference library.
• Tested the algorithm on 32 reference pigments and 10 Byzantine hagiography pigments, including binary mixtures.

Main Results:

• Successfully built a reference library of 32 pigments.
• The algorithm demonstrated good identification accuracy for unknown spectra.
• Effective identification was achieved despite the presence of noise, extra peaks, and background signals in the spectra.

Conclusions:

• Fuzzy logic provides an effective approach for the automated interpretation of Raman spectra of art pigments.
• The developed algorithm shows promise for non-destructive analysis and identification of materials in cultural heritage.
• This automated method can assist conservators and art historians in material identification and provenance studies.