Integrating Photogrammetric 3D City Models into Virtual Reality: A Methodological Approach Using Unreal Engine

Abstract. Three-dimensional (3D) city models have become indispensable in urban planning, disaster management, and various geospatial analyses. Advances in photogrammetric techniques and virtual reality (VR) platforms now enable highly detailed, immersive representations of urban environments for more effective decision-making. This paper presents a methodological framework for integrating produced photogrammetric 3D building models into Unreal Engine, with the goal of creating a VR environment. The workflow covers the capture of nadir and oblique aerial imagery, the organization of photogrammetric blocks, triangulation, and stereo compilation to derive accurate roof and facade geometries. Texture mapping is then applied using nadir and oblique photographs to ensure visual realism. Models in OBJ format are imported into Blender for scaling and preparation, and subsequently exported to FBX for compatibility with Unreal Engine. The paper also discusses incorporating georeferenced data via the Cesium for Unreal plugin. Despite a currently unresolved alignment issue that prevents proper placement of the models within Cesium’s global coordinate system, the outlined workflow provides a foundation for future research. Future improvements could include using pixel streaming for group VR sessions, adding GIS features to look up detailed information, and using real-time data for live updates. This method can help create interactive city simulations, making it easier for stakeholders to plan and analyze.