Tra Intelligenza Artificiale e H-BIM per la descrizione semantica dei beni culturali: la Certosa di Pisa. Artificial Intelligence and H-BIM for the Semantic Description of Cultural Heritage: the Pisa Charterhouse

Artificial Intelligence (AI) applications to architectural heritage usher promising evolutions in the Graphic arts and design field, as increasingly automated segmentation techniques can be applied to digital and numeric models for the semantic interpretation of survey data and subsequent recon-struction of conceptual and parametric 3D representations. Increasingly automated systems for the semantic description and classification of 3D data are indeed in ever-growing demand, also in view of optimizing Scan-to-BIM processes. In this context, the present contribution describes a semi-au-tomated methodological approach aimed at building semantically rich and intelligible representations starting from survey data. This approach features two core foundations: firstly, the classification and subsequent propagation of architectural types, by means of Artificial Intelligence (AI) algorithms applied to the point clouds; secondly, the reconstruction, in Building Information Modelling (BIM) environment, of the detected element classes, via information transmission procedures and visual programming languages. The analysis of the proposed methodology applies to representative case studies referring to selected settings of the Pisa Charterhouse.