Cultural-heritage and environmental diagnostics through ground-penetrating radar and artificial intelligence techniques

Ground-Penetrating Radar (GPR) is an instrument that, without requiring excavations or drilling, uses radar principles to obtain detailed information about the subsurface in a non-invasive manner. Moreover, the application of machine learning (ML) techniques to radar images (radargrams) makes it possible to learn or retrieve information that would otherwise be lost, allowing the most accurate predictions possible. Finally, these techniques can also be applied to material culture such as lithic industry, ceramics, or bone tools of even prehistoric peoples, allowing among other uses to predict the original dimensions of artefacts that have not reached us intact. Machine learning applied to cultural heritage is therefore an innovative and complementary tool for archaeology.