A Data Processing Chain for Terahertz Imaging and Its Use in Artwork Diagnostics

The effectiveness of terahertz (THz) imaging is dependent not only on the performances of the adopted hardware technology but also on the data processing approaches adopted by the users to elaborate the measured waveforms and obtain from them clear images of the object under test. With respect to data processing, this paper proposes a strategy involving three different steps aimed at reducing noise, filtering out undesired signal introduced by measurement system, and performing surface topography correction. The usefulness of the proposed data processing chain is preliminarily assessed by using data collected on a sample ad hoc prepared in laboratory. Afterward, an ancient mortar specimen, which is decorated by colored stucco, is analyzed by means of the proposed strategy.