A Digital Approach for the Analysis of Cultural Heritage – DIGITALIS

Data employed by archaeologists for analyzing and understanding the history of building, sites or artefacts are spread by their nature (terrain investigations, geological or physico-chemical measurements, iconographic or epigraphic studies, etc.), their level of reliability (observations, measurements, interpretations, assumptions, etc.), their nature (graphical, photography, textual, digital, etc.), and their status (existing or under elaboration, public or restricted access, etc.). Organizing and processing this data represents a difficult challenge for historians, who proceed intuitively and empirically, but also for computer scientists who labor for managing simultaneously complex and heterogeneous data.

Hence, using digital tools remains dedicated to storing and saving information, for its organization and occasional processing, but rarely as a support for thinking, reasoning, or producing scientific deductions. From two collections (ceramic and photograph libraries) and three patrimonial sites (Hypogée des dunes, Poitiers’ ducs palace, and Saint Savint’s abbey), the researchers of this consortium aim at overcoming the state of the art, for automating some deduction processes thanks to data inference (tracking contradictions and correspondences) and supporting the scientific reasoning.

Thanks to a close collaboration at all project stages, the researchers involved in this project aim at devising a digital system that integrates diversified information, associated with a hierarchical representation of annotations and a high-quality, photorealistic 3D visualisation tool based on actual data. From a computer science point of view, this project will rely on a structured data model, linked with an effective acquisition system which exploits machine learning for virtual restoration.