Machine Learning in Architectural Surveying: Possibility or Next Step of Development? From Photogrammetry to Augmented Reality of a Sculptural Group

In the practice of digital architectural surveying, as in traditional architectural surveying, in-situ data acquisition operations are the components perceived as most structuring for scholars in other subjects and fields of research. In the subject, however, it is the data processing stage that is the most significant component of the work path, forming the core of the process. Connected to it, the restitution stage is necessary to find strategies for communicating the synthesis of survey operations. In the digitization of Cultural Heritage, the digital component is the means of connection between the processes of data processing and their purpose, summarized by the chosen method of restitution. In the panorama of digital surveying techniques, artificial intelligence and machine learning do not yet find significant application. Only with the development of tools for acquiring the built environment will it be possible to find a real use of automated assessment and independence from the operator. Acquisition of the data, it is the processing and return phases that constitute the most fertile scenario for the development of artificial intelligence. The development of machine learning toward knowledge of geometric feature recognition or digital model optimization functional to the restitution method can be two important strategies for obtaining results characterized by a high level of automation.