COMBINING VISIBILITY ANALYSIS AND DEEP LEARNING FOR REFINEMENT OF SEMANTIC 3D BUILDING MODELS BY CONFLICT CLASSIFICATION

Abstract. Semantic 3D building models are widely available and used in numerous applications. Such display rich semantics but no façade openings, chiefly owing to their aerial acquisition techniques. Hence, refining models’ façades using dense, street-level, terrestrial point clouds seems a promising strategy. In this paper, we propose method of combining visibility analysis neural networks for enriching with window door features. the method, occupancy voxels fused classified clouds, which provides voxels. Voxels also identify conflicts between laser observations models. The semantic combined Bayesian network classify delineate reconstructed model library. Unaffected is preserved while updated one added, thereby upgrading LoD3. Moreover, results back-projected onto improve points’ classification accuracy. We tested our on municipal CityGML LoD2 repository open cloud datasets: TUM-MLS-2016 TUM-FAÇADE. Validation revealed that improves accuracy segmentation upgrades buildings elements. can be applied enhance urban simulations facilitate development algorithms.