Multispectral UV fluorescence analysis of painted surfaces

0
301

The article “Multispectral UV Fluorescence Analysis of Painted Surfaces” by Anna Pelagotti, L. Pezzati, Alessandro Piva, and Andrea Del Mastio, published in 2006, explores the utility of a novel system for acquiring ultraviolet (UV)-induced fluorescence images on painted surfaces. This work brings significant technological and methodological advances that improve the precision and applicability of the technique in the field of art conservation.

Since its inception in the 1920s, observing artworks under UV light has been a technique primarily used to identify old varnish layers and retouched or repainted areas. However, the capability of this traditional technique has been limited by the lack of precision in colorimetric reproduction and radiometric evaluation of fluorescence images, resulting in merely qualitative outcomes. To overcome these limitations, the article presents an advanced UV system that enables the digital capture and processing of multispectral fluorescence images. This system includes a specifically designed UV light source, a scientific CCD camera, and interference filters, all controlled through multivariate analysis software.

The developed system allows for correction and calibration of the captured images, as well as the recombination of seven monochromatic bands into an RGB image that emulates how the human eye would perceive the painting under UV light. This enables the separation of fluorescence components and reflectance components caused by scattered visible light. In this way, the use of spectral profiles to identify pigments and binding materials in paintings is facilitated.

Methodologically, the image processing is complex and encompasses several crucial steps. First, image registration is performed to correct geometric misalignments caused by differences in the optical paths of various filters. Subsequently, images are corrected for uneven illumination and scattered light, and decomposed into fluorescence and reflectance components. The final process involves radiance calibration carried out through indirect procedures using reflectance standards and spectroradiometers. Through these methods, the system ensures flexibility to translate complex spectral measurements into meaningful radiometric and photometric data.

The authors also present practical applications of their system. Using the algorithm known as Spectral Angle Mapper (SAM), the system was employed to identify different pigments in specific details of a painting, showing promising results by combining fluorescence and reflectance images. It was demonstrated that the best pigment identifications are achieved by combining both reflectance and fluorescence images, especially to distinguish restored areas from original parts of the artwork.

In conclusion, the article presents a significant improvement for traditional fluorescence imaging techniques in art conservation by providing precise and reproducible quantitative data. The capability of this system to separate and analyze spectral components opens new dimensions in the diagnosis and conservation of artworks, overcoming the limitations of previous approaches and establishing a new standard in the field.

Review on the publisher’s website