Agile digitization for historic architecture using 360° capture, deep learning, and virtual reality

Abstract

The agile digitization of historic buildings is becoming increasingly critical for preservation, conservation, and maintenance in response to climate change, geopolitical conflicts, and other threats of destruction. This paper explores whether deep learning-based novel-view synthesis, combined with commercial 360° cameras and standalone virtual reality headsets, can streamline the digitization process for historic architecture. A case study of a historic interior in Québec, Canada, is used to evaluate the method's capacity to enhance agility, accuracy, and efficiency. The findings demonstrate that this approach significantly reduces complexity, labor, cost, and time while improving precision and workflow. These outcomes offer particular value to heritage experts, building engineers, and creative professionals seeking practical tools for agile digitization of historic architecture. By advancing digitization methods, this study also inspires future research into the broader applications of deep learning and immersive technologies for cultural heritage preservation.