Generation of LOD4 models for buildings towards the automated 3D modeling of BIMs and digital twins
Bryan German Pantoja-Rosero, Alexander Rusnak, Frédé́ric Kaplan, Katrin Beyer
Abstract
An image-based methodology is presented for the automatic generation of geometric building models at LOD4, incorporating both interior and exterior geometrical information. Existing approaches often focus on simplified geometries for either exteriors or interiors, leading to integration challenges due to data complexity and processing demands. This methodology addresses these challenges by utilizing three structure-from-motion models: one for the building exterior, one for the interior, and one for the entrance. The exterior and interior data are processed separately using planar primitives, and the models are subsequently aligned through a 3D point cloud registration method based on 2D image features. This ensures a unified coordinate system and accurate generation of the LOD4 model. The framework achieved a mean relative error of 3.06% and a mean absolute error of 0.05 m, underscoring its robustness for applications such as numerical modeling, construction management, and structural health monitoring, making it valuable for further advancements in building information models and digital twins. • Framework for automatically reconstructing LOD4 models of free-standing buildings. • Simplified methodology for generation of interior LOD (ILOD0) using unstructured point clouds. • Use of a 3D registration method using 2D image features to combine building interiors and exteriors. • Benchmark dataset for evaluating the generation of LOD4 models from point clouds.