Romano-Egyptian Mummy Portraits from Tebtunis, Egypt
The Hearst portraits were studied using non-destructive and non-invasive imaging techniques that extract information about their underlying shape and color. Using a computational imaging algorithm called photometric stereo, quantitative measurements were made of brush and tool marks. Superposition of different layers was also determined to establish the order of painting of the various elements of decoration. Hyper-spectral imaging data was collected between the UV through NIR for measurements of color. These color data were mined using machine-learning algorithms to identify pigments on a pixel by pixel basis. Using this information as a guide, micro-samples of paint were removed from discrete areas and prepared as cross-sections for a ground truth determination of the colorants used. These cross-sections were analyzed using a combination of Raman and scanning electron microscopy which revealed a rich palette of both organic and inorganic pigments that included a variety of iron earths, red lead, as well as indigo and Egyptian blue. The specific information obtained about painting methods and the overall distribution of pigments produces groupings of portraits made of similar materials and has led to the identification of ancient workshops and the hand of specific artists.