Ai-Based 3D Reconstruction of Villino Rutelli from a Single Archival Photograph

Abstract

This paper analyses a three-dimensional reconstruction process based on the implementation of Arti¬ficial Intelligence, utilizing a single historical image as input. Hence, the different 3D models were developed using archival photographs and metric data drawn from historical documentation. The research is situated within the broader field of studies on urban transformations in Palermo, particularly focusing on the reconstruction of buildings and villas from the late 19th and early 20th centuries, which have since disappeared but once shaped the city’s architectural identity. The case study focuses on Villino Rutelli, a neo-Gothic building that was demolished during a major phase of urban expansion to accommodate emerging needs and new residential typologies. The models generated through AI-based processes were qualitatively evaluated through comparison with iconographic sources and previous reconstruction hypotheses developed in academic contexts. The results represent an initial experimental phase, highlighting the potential of innovative technologies for the digital restitution of lost archi¬tecture and, more broadly, for the valorisation of urban cultural heritage.