Abstract. Modelling cultural heritage necessitates the use of specific software and file formats that can be difficult to transpose between one another. These interoperability challenges are particularly acute when employing parametric modelling tools, whether within the context of HBIM or 4D modelling (i.e., the reconstruction of past states of heritage objects). Standard export workflows often fail to preserve complex procedural textures or require time-consuming manual UV unwrapping, rendering direct conversion inefficient. In this study, the potential of applying photogrammetry to synthetic images is investigated as a solution to these limitations. A rigorous dataset is generated in Blender based on a 4D parametric model of a Rhenish castle currently in ruins, utilising path tracing to ensure visual fidelity. This method ensures interoperability by consolidating multiple complex, software-dependent parametric objects into a single, universally compatible mesh. This transformation allows the model to be directly employed across diverse software environments, independent of the original creation tools. Subsequently, the photogrammetric mesh is decimated and optimised to facilitate sharing and viewing on all computer configurations. To validate the approach, a quantitative evaluation is performed: the geometric deviation between the original parametric model and the reconstructed mesh is analysed, and rendering performance is benchmarked to demonstrate the method’s suitability for dissemination and visualisation tools.
Sommer et al. (2026) studied this question.