RouteExtract: A Modular Pipeline for Extracting Routes from Paper Maps

2025-09-15

AI Summary

This work addresses the challenge of directly utilizing scanned paper maps for GPS navigation. We propose an end-to-end trail network extraction framework: first, georegistering scanned maps to align them spatially; second, applying a U-Net architecture for robust binary segmentation of trails; third, constructing a graph representation to encode topological relationships, and iteratively refining it via a routing-engine-driven optimization that leverages real-world path reachability feedback. This modular design significantly enhances adaptability and generalization across diverse paper map types—including hand-drawn, printed, and faded variants. Experiments demonstrate that our method consistently reconstructs complex trail networks and generates high-precision GPS-compatible paths satisfying real-world navigation requirements. Quantitatively, it outperforms conventional pipelines in both topological completeness and geometric accuracy.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Intelligent Robots: Localization, Mapping, and Navigation
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding
• Web Mining and Content Analysis: Topic discovery and tracking

Abstract

Paper maps remain widely used for hiking and sightseeing because they contain curated trails and locally relevant annotations that are often missing from digital navigation applications such as Google Maps. We propose a pipeline to extract navigable trails from scanned maps, enabling their use in GPS-based navigation. Our method combines georeferencing, U-Net-based binary segmentation, graph construction, and an iterative refinement procedure using a routing engine. We evaluate the full end-to-end pipeline as well as individual components, showing that the approach can robustly recover trail networks from diverse map styles and generate GPS routes suitable for practical use.

Problem

• Extracting navigable trails from scanned paper maps
• Enabling paper map trails for GPS navigation use
• Robustly recovering trail networks from diverse map styles

Innovation

• U-Net-based binary segmentation for trails
• Graph construction from extracted map features
• Iterative refinement using routing engine