CM2LoD3: Reconstructing LoD3 Building Models Using Semantic Conflict Maps

2025-08-21

AI Summary

To address the low automation in reconstructing facade details (e.g., windows, doors, underground passages) for large-scale LoD3 building models, this paper proposes a reconstruction framework based on a semantic conflict graph. The method introduces a Semantic Conflict Graph Generator (SCMG) that integrates ray-model prior analysis with synthetic and real-world data to construct conflict graphs encoding geometric–semantic inconsistencies. Furthermore, it incorporates an uncertainty-aware multimodal fusion mechanism that jointly leverages semantic segmentation outputs and confidence-weighted texture segmentation results to improve opening region detection accuracy. Experiments demonstrate significant improvements in facade opening segmentation and 3D reconstruction performance: IoU reaches 61% when texture information is fused. This work establishes a novel paradigm for high-fidelity, scalable urban modeling—critical for digital twin systems and smart city applications.

Technology Category

• Computer Vision: Segmentation
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Detailed 3D building models are crucial for urban planning, digital twins, and disaster management applications. While Level of Detail 1 (LoD)1 and LoD2 building models are widely available, they lack detailed facade elements essential for advanced urban analysis. In contrast, LoD3 models address this limitation by incorporating facade elements such as windows, doors, and underpasses. However, their generation has traditionally required manual modeling, making large-scale adoption challenging. In this contribution, CM2LoD3, we present a novel method for reconstructing LoD3 building models leveraging Conflict Maps (CMs) obtained from ray-to-model-prior analysis. Unlike previous works, we concentrate on semantically segmenting real-world CMs with synthetically generated CMs from our developed Semantic Conflict Map Generator (SCMG). We also observe that additional segmentation of textured models can be fused with CMs using confidence scores to further increase segmentation performance and thus increase 3D reconstruction accuracy. Experimental results demonstrate the effectiveness of our CM2LoD3 method in segmenting and reconstructing building openings, with the 61% performance with uncertainty-aware fusion of segmented building textures. This research contributes to the advancement of automated LoD3 model reconstruction, paving the way for scalable and efficient 3D city modeling. Our project is available: https://github.com/InFraHank/CM2LoD3

Problem

• Automating LoD3 building model reconstruction with facade details
• Semantically segmenting conflict maps for 3D building openings
• Fusing texture segmentation with conflict maps to improve accuracy

Innovation

• Semantic Conflict Maps for LoD3 reconstruction
• Synthetic and real-world map fusion
• Uncertainty-aware texture fusion enhancement