DeH4R: A Decoupled and Hybrid Method for Road Network Graph Extraction

2025-08-19

AI Summary

Road network extraction from remote sensing imagery faces a fundamental trade-off between topological fidelity and inference efficiency: segmentation-based methods suffer from topological fragmentation after vectorization; graph-growing approaches ensure topological accuracy but incur high computational cost; while graph-generation methods are efficient, they lack dynamic node insertion capability. This paper proposes a decoupled hybrid architecture that decomposes the task into four sequential stages—candidate vertex detection, adjacency prediction, initial graph construction, and iterative graph expansion—thereby synergistically integrating the efficiency of graph generation with the topological adaptability of graph growing. The design enables runtime dynamic node insertion, achieving both high-fidelity topology and spatial consistency. Evaluated on CityScale and SpaceNet, our method achieves state-of-the-art performance: +4.62 in APLS, +10.18 in IoU, and ~10× faster inference than RNGDet++.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Machine Learning: Graph-based Machine Learning
• Computer Vision: Segmentation

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines

Abstract

The automated extraction of complete and precise road network graphs from remote sensing imagery remains a critical challenge in geospatial computer vision. Segmentation-based approaches, while effective in pixel-level recognition, struggle to maintain topology fidelity after vectorization postprocessing. Graph-growing methods build more topologically faithful graphs but suffer from computationally prohibitive iterative ROI cropping. Graph-generating methods first predict global static candidate road network vertices, and then infer possible edges between vertices. They achieve fast topology-aware inference, but limits the dynamic insertion of vertices. To address these challenges, we propose DeH4R, a novel hybrid model that combines graph-generating efficiency and graph-growing dynamics. This is achieved by decoupling the task into candidate vertex detection, adjacent vertex prediction, initial graph contruction, and graph expansion. This architectural innovation enables dynamic vertex (edge) insertions while retaining fast inference speed and enhancing both topology fidelity and spatial consistency. Comprehensive evaluations on CityScale and SpaceNet benchmarks demonstrate state-of-the-art (SOTA) performance. DeH4R outperforms the prior SOTA graph-growing method RNGDet++ by 4.62 APLS and 10.18 IoU on CityScale, while being approximately 10 $ imes$ faster. The code will be made publicly available at https://github.com/7777777FAN/DeH4R.

Problem

• Extracting complete road networks from remote sensing imagery
• Maintaining topology fidelity after vectorization postprocessing
• Balancing computational efficiency with dynamic vertex insertion

Innovation

• Hybrid model combining graph-generating efficiency and graph-growing dynamics
• Decouples task into vertex detection, adjacency prediction, and graph expansion
• Enables dynamic vertex insertions while maintaining fast inference speed