Rectifying the emissivity variations problem caused by pigments in artworks inspected by infrared thermography: A simple, useful, effective, and optimized approach for the cultural heritage field

Abstract

Inspection of cultural heritage objects plays nowadays paramount importance around the world. Non-destructive inspection is a must and a necessity in order to preserve the integrity of the artwork without losing any precious material composing it. The use of thermal non-destructive inspection is a good idea since, by exploiting the 3D diffusion inside the object, triggered by external radiation, surface and subsurface defects may be revealed. To do this, the long-wave infrared (LWIR) spectrum is usually exploited in combination with a thermal camera. In the cultural heritage field, the main problem to solve to detect as much as possible thermal imprints linked to invisible defects is the minimisation of the impact of emissivity variations caused by pigments composing the colours. A simple, effective, solid, and optimized method for decorated paintings is here applied right after some preliminary results. It is based on active thermography as the modality of inspection, and a thermal stimulus provoked by halogen lamps; thermal images recorded on a panel painting including man-made defects have been analysed and processed in MATLAB® environment. After extraction of de-nosing functions based on the heating and cooling steps, two methods are proposed to enhance the de-noised thermograms. Then, popular methods of Pulsed Phase Thermography (PPT) and the Principal Component Thermography (PCT) are applied. In the end, after minimizing the effect of emissivity variation, an optimization fusion proposal through post-processing the emissivity adjusted thermograms is provided. A brief but exhaustive review introduce and guide the readers towards the problem for which the authors took a step ahead via a proposal based on image fusion.