Publication Details
Abstract
This article examines the scientific and methodological foundations of applying artificial intelligence (AI) technologies to digital documentation, 3D modelling and virtual reconstruction of historical architectural heritage. The study considers photogrammetry, terrestrial laser scanning, UAV survey, computer vision, neural networks, NeRF, generative models and HBIM as an interconnected technological chain for heritage conservation. The key argument of the paper is that a 3D model of a historical monument should not be treated merely as a visual representation, but as a digital scientific document integrating geometry, material evidence, historical interpretation, structural condition and restoration metadata. As a result, the paper proposes a stage-based methodological framework, validation criteria and ethical limitations for AI-assisted modelling of architectural heritage. The findings are relevant for the development of digital monitoring, virtual museums, educational platforms and scientific conservation projects in historic cities of Uzbekistan such as Samarkand, Bukhara, Khiva and Shakhrisabz.