Segment Anything Model for Road Network Graph Extraction
Preprint
Abstract
We propose SAM-Road, an adaptation of the Segment Anything Model (SAM) for extracting large-scale, vectorized road network graphs from satellite imagery. To predict graph geometry, we formulate it as a dense semantic segmentation task, leveraging the inherent strengths of SAM. The image encoder of SAM is fine-tuned to produce probability masks for roads and intersections, from which the graph vertices are extracted via simple non-maximum suppression. To predict graph topology, we designed a lightweight transformer-based graph neural network, which leverages the SAM image embeddings to estimate the edge existence probabilities between vertices. Our approach directly predicts the graph vertices and edges for large regions without expensive and complex post-processing heuristics, and is capable of building complete road network graphs spanning multiple square kilometers in a matter of seconds. With its simple, straightforward, and minimalist design, SAM-Road achieves comparable accuracy with the state-of-the-art method RNGDet++, while being 40 times faster on the City-scale dataset. We thus demonstrate the power of a foundational vision model when applied to a graph learning task. The code is available at https://github.com/htcr/sam_road.Keywords
All Related Versions
This publication has 15 references indexed in Scilit:
- GPT-4V(ision) is a Human-Aligned Evaluator for Text-to-3D Generation2024
- RNGDet++: Road Network Graph Detection by Transformer With Instance Segmentation and Multi-Scale Features EnhancementIEEE Robotics and Automation Letters, 2023
- Exploring Plain Vision Transformer Backbones for Object Detection2022
- TD-Road: Top-Down Road Network Extraction with Holistic Graph Construction2022
- RNGDet: Road Network Graph Detection by Transformer in Aerial ImagesIEEE Transactions on Geoscience and Remote Sensing, 2022
- Swin Transformer: Hierarchical Vision Transformer using Shifted Windows2021
- HDMapGen: A Hierarchical Graph Generative Model of High Definition Maps2021
- A compact review of molecular property prediction with graph neural networksDrug Discovery Today: Technologies, 2020
- Towards Segmenting Anything That Moves2019
- RoadTracer: Automatic Extraction of Road Networks from Aerial Images2018