BigEarthNet v1 (BigEarthNet-MM): Sentinel-1 and Sentinel-2 Multi-Label Land Cover Patches with CORINE 2018 Labels
- Provider
- BigEarthNet team, TU Berlin (Sumbul et al. 2019)
- License
- CDLA-Permissive-1.0 (source)
- Format
- JSON (in sampled files)
- Stored at
- Azure Blob Storage, https://radiantearth.blob.core.windows.net/mlhub/bigearthnet
- Access
- Free and public. No account with the data provider is needed.
- Category
- Earth Observation & Geospatial
- Source
- https://bigearth.net/
About this dataset
The original BigEarthNet benchmark (version 1) for multi-label land cover classification of satellite image patches, as hosted on Radiant MLHub. 590,326 Sentinel-2 Level-2A patches were cut from 125 tiles acquired between June 2017 and May 2018 over Austria, Belgium, Finland, Ireland, Kosovo, Lithuania, Luxembourg, Portugal, Serbia and Switzerland, atmospherically corrected with sen2cor, and labeled with multiple CORINE Land Cover 2018 (CLC 2018) classes; the matching Sentinel-1 GRD patches (BigEarthNet-S1, VV and VH backscatter) are included, making the multi-modal BigEarthNet-MM set. Each patch is a folder named after its source scene and grid position (S2A_MSIL2A_..._row_col/ or S1A_IW_GRDH_..._row_col/) holding one GeoTIFF per band and a labels_metadata.json with the CLC labels and acquisition details. Use for multi-label scene classification, self-supervised pretraining and SAR-optical fusion with PyTorch, TorchGeo or rasterio. This is v1; the reprocessed BigEarthNet v2.0 (reBEN) is a separate catalog row. Licensed CDLA-Permissive-1.0; cite Sumbul, Charfuelan, Demir and Markl (2019), IGARSS, pp. 5901-5904.
How to access this dataset in Python
import adlfs
fs = adlfs.AzureBlobFileSystem(account_name='radiantearth', anon=True)
fs.ls('mlhub/bigearthnet/')[:10]Lists the files with anonymous access. Runs on any machine, with no cloud account.
Data structure
File types in sampled files: JSON
Files
Example file: .source/metadata.json
Analyze this dataset without downloading it
On Scigantic, BigEarthNet v1 (BigEarthNet-MM): Sentinel-1 and Sentinel-2 Multi-Label Land Cover Patches with CORINE 2018 Labels opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.
This page is also available as Markdown and as JSON.