Hong Kong Road Surface Data with AI-Derived Paved and Unpaved Classifications

OpenStreetMap road data for Hong Kong is augmented with deep learning predictions from Mapillary imagery to classify surfaces as paved or unpaved. The dataset, created by HeiGIT, was last updated on March 2, 2026. It includes OSM attributes and urban classification layers from GHSU and AFRICAPOLIS.

Use Cases

• Transportation planning based on road surface type and urban classification attributes.
• Infrastructure analysis based on OSM attributes like highway type, width, and lanes.
• Climate emissions assessment based on the proportion of paved versus unpaved roads.
• GIS applications for routing and logistics based on combined surface classifications.

Strengths

• Includes AI-derived surface classifications from a hybrid deep learning approach detailed in a referenced arXiv paper.
• Combines multiple data sources: OpenStreetMap, Mapillary imagery, GHSU, and AFRICAPOLIS urban layers.
• Provides multiple classification fields (OSM, DL-prioritized, OSM-prioritized) for comparison and validation.

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.
• Key statistics in the description show placeholder values (0.0 million km, nan%), indicating potential data gaps.

Provenance

Source

OpenStreetMap (OSM), augmented with Mapillary imagery and predictions by HeiGIT.

Collection Method

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

Time Range

Mean timestamp of predictions is included (DL_mean_timestamp), but specific range is not stated.

Freshness

Last updated 2026-03-02 17:08:29.551051; freshness should be verified.

Geography

China, Hong Kong Special Administrative Region.