Speaker
Description
The analysis of Cultural Heritage samples imposes some requirements during the development of the analytical methods that would be applied to investigate them. In this regard, non-invasiveness is often key to ensure the conservation of the sample, or at least to imply minimal alteration. This contribution presents a digital image-based method to distinguish among different inorganic pigments in mural paintings. Different laboratory mock-ups were prepared and painted with 12 different inorganic and historical pigments, covering four hues: red, green, blue and yellow. Each hue was represented by three different pigments. Thereafter, the pigments were colorimetrically characterised using digital image colorimetry, and different chemometric classification methods were applied. Results demonstrated that Support Vector Machines allowed to identify the specific pigment with errors below 5.7 %. Thus, this smartphone-based method permits to obtain knowledge about the composition of the samples with no direct contact with the sample, and provides objective information that does not rely on subjective interpretation or environmental factors. Also, it emerges as an affordable and easy-to-use tool for restorers and researchers.
| Presenting author | Roberto Sáez-Hernández |
|---|
