---
dataset_info:
  features:
  - name: rgb
    dtype: image
  - name: bands
    dtype: binary
  - name: bands_shape
    list: int64
  - name: bands_dtype
    dtype: string
  - name: band_metadata
    dtype: string
  - name: band_order
    list: string
  - name: label
    dtype:
      class_label:
        names:
          '0': AnnualCrop
          '1': Forest
          '2': HerbaceousVegetation
          '3': Highway
          '4': Industrial
          '5': Pasture
          '6': PermanentCrop
          '7': Residential
          '8': River
          '9': SeaLake
  splits:
  - name: train
    num_bytes: 3001912034
    num_examples: 27000
  download_size: 2661440828
  dataset_size: 3001912034
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
license: mit
task_categories:
- image-classification
size_categories:
- 10K<n<100K
---
# Eurosat Lulc Classification

This dataset provides multispectral satellite imagery capturing diverse land use and land cover patterns across European landscapes. The real-world images, collected from satellite platforms, offer a comprehensive resource for training and evaluating computer vision models in environmental monitoring and land cover analysis. The dataset contains 27,000 images across 10 classes: AnnualCrop, Forest, HerbaceousVegetation, Highway, Industrial, Pasture, PermanentCrop, Residential, River, SeaLake.  
Images per class:
- AnnualCrop: 3,000
- Forest: 3,000
- HerbaceousVegetation: 3,000
- Highway: 2,500
- Industrial: 2,500
- Pasture: 2,000
- PermanentCrop: 2,500
- Residential: 3,000
- River: 2,500
- SeaLake: 3,000

This dataset is indexed on https://project-agml.github.io/ as part of the AgML python library.

## Citation

```bibtex
@article{helber2019eurosat,
  title={Eurosat: A novel dataset and deep learning benchmark for land use and land cover classification},
  author={Helber, Patrick and Bischke, Benjamin and Dengel, Andreas and Borth, Damian},
  journal={IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing},
  volume={12},
  number={7},
  pages={2217--2226},
  year={2019},
  publisher={IEEE}
}
```

Helber, P., Bischke, B., Dengel, A., & Borth, D. (2018). EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classification [Dataset]. In EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classification (Vol. 12, Issue 7, pp. 2217–2226). Zenodo. Introducing Eurosat: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classification, Valencia, Spain. https://doi.org/10.5281/zenodo.7711810

Sentinel data is open source. Please refer to https://sentinels.copernicus.eu/documents/247904/690755/Sentinel_Data_Legal_Notice for terms of use. 

*This dataset was reformatted from its original format to match HuggingFace standards.*