Road Surface Classification for South Korea from OSM and Satellite Imagery

Approximately 471,300 km of roads are mapped in OpenStreetMap for the Republic of Korea, with AI-derived estimates classifying 75,100 km as paved and 31,800 km as unpaved. The dataset, created by the Heidelberg Institute for Geoinformation Technology (HeiGIT), combines OSM data with deep learning predictions from Mapillary imagery and urban layers. It was last updated on March 2, 2026.

Use Cases

• Transportation planning based on paved versus unpaved road classifications.
• Infrastructure analysis based on OSM attributes like highway type, surface, and width.
• Climate emissions assessment based on road surface material and urban/rural classification.
• GIS applications for routing and logistics based on the geospatial road network.

Strengths

• Provides AI-derived surface classification for 471,300 km of roads.
• Includes multiple classification schemes (OSM-priority and DL-priority) for comparison.
• Augments OSM data with deep learning predictions to address missing surface information for 364,500 km of roads.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.

Provenance

Source

OpenStreetMap (OSM), Mapillary imagery, GHSU Global Human Settlement Urban Layer 2019, AFRICAPOLIS 2020 urban layer.

Collection Method

Surface classification derived using a hybrid deep learning approach on street-level imagery.

Time Range

Mean timestamp of predictions is provided in the 'DL_mean_timestamp' field.

Freshness

Last updated 2026-03-02 17:21:04.499245; freshness should be verified.

Geography

Republic of Korea (South Korea).