Multispectral imaging of paintings: A way to material identification

In this article, we propose a novel multispectral digital imaging technique, concurrently analyzing different imaging modalities with rather simple hardware as a reliable way to achieve material localization and identification on the entire painting surface at one time. First, we compare the currently available systems, as far as both imaging and spot methods are concerned. We then describe both our acquisition device and the employed signal processing algorithms. Some interesting results are presented, showing the system performance both on lab samples and on an original painting from the fourteenth century.