Abstract
This study investigates the application of NDTs for the detection of sub-surface defects in an ancient book, with the aim of improving conservation methods in the field of cultural heritage. A sequence of thermographic images in a solar loading thermography (SLT) scenario was acquired during a diagnostic campaign in Harbin, China, to identify four distinct fabricated dowels made of Wool, Rubber, Teflon(®), and Synthetic material. The images were processed in two ways: the first combined advanced image-processing methods: pre-processing via MdFIF, post-processing, PCT and RPCT, applied both to the original sequence and to the MdFIF-filtered thermograms. The second approach employed numerical simulations in COMSOL Multiphysics(®) to develop a predictive thermal model. The comparison of localized thermal anomalies obtained from the two approaches demonstrated the capability of NDTs to reliably reveal artificial defects, confirming their suitability for diagnostic conservation. Overall, the integration of advanced image processing with numerical simulation enhances diagnostic accuracy, particularly for subtle or low-contrast anomalies, thereby enabling more informed condition assessment and supporting rapid, targeted, and preventive conservation strategies.