dataset · Global

LandCover.ai

Hugging Face dataset dragon7/LandCover.ai

Feature or claim this listing

Represent LandCover.ai? Send a claim for review so we can discuss listing updates. Claims do not automatically grant ownership or verification.

Data offered docs-verified: RGB orthophotos (aerial imagery) · land-cover masks: building, woodland, water, road
Coverage docs-verified: Poland, Central Europe
Resolution docs-verified: 25 cm per pixel (33 orthophotos); 50 cm per pixel (8 orthophotos)
Formats docs-verified: GeoTIFF
Access docs-verified: Public Hugging Face dataset; licence CC BY-NC-SA 4.0 (non-commercial, share-alike); no auth in ENTRY

Use cases llm-inferred: Non-commercial R&D prototype of building/woodland/water/road segmentation on aerial imagery · Benchmarking segmentation models for habitat-type mapping methods before site-specific training · Testing GeoTIFF raster tiling and masking pipelines in EPSG:2180
Audience llm-inferred: GIS analysts · remote-sensing engineers · ML engineers · ecologists (method development only)

Constraints llm-inferred: CC BY-NC-SA 4.0 non-commercial clause likely excludes paid consultancy use; confirm with licensor · Coverage is Poland only; the ENTRY region tag 'Global' is not supported by the docs · RGB only (three spectral bands); no NIR or LiDAR · Only four classes; roads (3.5 km2) and buildings (1.85 km2) are sparse · Mixed resolution: 8 of 41 orthophotos are at 50 cm · No update or maintenance statement; version 1 tied to a 2020 paper

Verified 2026-10-15 via HF API: exists; 542 downloads/30d, 0 likes.

Metadata observed 14/10/2026 · research sweep: manual-2026-10-15 (Hugging Face datasets)