Processing of UV/VIS/nIR/mIR diffuse reflectance spectra and acoustic microscopy echo graphs for stratigraphy determination, using neural networks and wavelet transform

A main concern for the documentation of artworks (such as paintings, miniatures, icons, manuscripts, wall paintings) is how to reveal the information that stems from the internal paint layers (stratigraphy). This information is up to now mainly acquired through analytical spectroscopic methods which require a micro-sampling operation. In the majority of the cases, the objects under study are extremely valuable, rare and fragile, therefore they may not be subjected to any intervention, rendering the micro sampling procedures prohibitive. The current work proposes a method combining signal processing techniques with novel non-destructive testing techniques for the determination of artworks´ paint layers profile in depth as well as the local distribution of the pigments in each of the paint layers without micro-sampling.