Multispectral imaging of papyri: area segregation by evaluation of their spectral signature correlation

摘要

Multispectral imaging technology was applied to deteriorated ancient manuscripts. Monochrome images were registered at twelve wavelengths. An image processing algorithm that utilises relative spectral intensity distribution for the areas of the surface (rather than absolute values of intensity of the reflected light) was evaluated. It was found that the areas in a cubed MSI image representing ink and papyrus were separable due to their distinctive spectral signatures. The developed technique for enhancing the contrast of the inscriptions was shown to work effectively over a large range of lightness for both ink and papyrus surface.