Automatic Extraction of Road Alignment and Geometric Features from Airborne Photogrammetric Point Clouds
Accurate and up-to-date information on roadway alignment and geometric parameters is essential for crash risk assessment, geometric design evaluation, and network-scale infrastructure analysis. We present a scalable workflow to automatically extract road alignment and geometric parameters from airborne photogrammetric point clouds. The method isolates the pavement surface through spectral-attribute filtering, density-based clustering, and parametric refinement then derives the road centerline and edges and computes horizontal and vertical alignment parameters. Validation against manually delineated references on 15 segments of 30 m shows average correctness of 84.10% and completeness of 83.94% at a 25 cm buffer, rising to 98.00% and 98.03% at 40 cm, and 99.94% and 99.97% at 50 cm, respectively, with point density of about 4 points per square meter ( pts / m 2 ). As a brief illustration of a practical application, we include a simple case study that uses the extracted parameters in a basic crash-risk model. The workflow enables network-scale extraction of alignment and geometric parameters from widely available airborne data.
Authors
- Álvaro García (ORCID: https://orcid.org/0000-0001-8730-6632)
- Stefan Biermeier
- Mohammadhossein Rezazadeh
Institutions
- RWTH Aachen University (DE)
Publication Details
- Journal
- Journal of Computing in Civil Engineering
- Published
- 2026-11-18
- DOI
- https://doi.org/10.1061/jccee5.cpeng-7855
- Primary Topic
- Automated Road and Building Extraction
- Type
- article
- Field-Weighted Citation Impact
- 0.00