Transportation
African Road Image Dataset
Street-scene schema for road condition labels. Images are not included.
Dataset description
Documents how a road dataset should be labelled: surface, shoulder, and obstruction. No photographs are stored, and no number plate or face data is accepted.
- Creator
- AI HUB Data Team (demo)
- Records
- 18,000
- Format
- JPG (not hosted)
- Collection period
- Not collected
- Updated
- 2026-11-07
- Downloads of samples
- 0
Quality score 85/100
Research sample
Samples use a 10-minute grant. Full files are not hosted.
Quality breakdown
20% completeness + 18% accuracy + 14% consistency + 12% annotation agreement + 14% documentation + 12% duplicate health + 10% metadata.
Completeness
80%
Accuracy
85%
Consistency
84%
Duplicate rate
1%
Annotation agreement
82%
Documentation
86%
Metadata
78%
Data preview
Safe sample only. Restricted fields are not shown.
| frame_id | surface | wet | obstruction |
|---|---|---|---|
| demo | unpaved | no | none |
Data card
- Purpose
- Road condition classification for maintenance research.
- Collection method
- Would require a privacy review before any capture of people or plates.
- Annotation method
- Surface class only. Faces and plates would be out of scope.
- Source
- Demo card.
- Size
- 18,000 catalogue target. Sample only is hosted.
- Fields
- frame_id, surface, wet, obstruction
- Language
- English
- Geography
- Nigeria · Ghana
- Limitations
- No imagery.
- Potential bias
- A real set must include unpaved roads, not only highways.
- Recommended uses
- Writing an annotation guide.
- Prohibited uses
- Surveillance, enforcement, or identifying road users.
- Licence
- research
Versions
v0.8 · 2026-03-05
Policy card. Capture not approved.
+0 / −0 · Privacy noted, files absent · surface, wet, obstruction
Report dataset
Use this for privacy, copyright, harmful content or a misleading card. Do not attach more personal data than the report needs.
You can also use the report desk.