qualia_core.dataset.EuroSAT module

## EuroSAT.py: A Qualia dataset for land use and land cover classification using Sentinel-2 satellite images.

## Author # - Jonathan Courtois # [jonathan.courtois@univ-cotedazur.fr](mailto:jonathan.courtois@univ-cotedazur.fr) ## Dataset Reference # - EuroSat Dataset: # https://github.com/phelber/EuroSAT # - Installation Instructions: # The EuroSat dataset .zip files must be uncompressed in your ‘dataset’ folder of the Qualia repository with this structure: # .dataset/ # ├── EuroSAT/ # ├── ├── MS/ # ├── ├── ├── [Class_folder]/ # ├── ├── ├── ├── [.tif] # ├── ├── RGB/ # ├── ├── ├── [Class_folder]/ # ├── ├── ├── ├── [.jpg] # - Citation: # @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}, # year={2019}, # publisher={IEEE}

class qualia_core.dataset.EuroSAT.EuroSAT(path: str = '', variant: str = 'MS', dtype: str = 'float32', train_test_ratio: float = 0.8)[source]

Bases: RawDataset

EuroSAT Land Use and Land Cover Classification with Sentinel-2,.

Challenge of land use and land cover classification using Sentinel-2 satellite images. The Sentinel-2 satellite images are openly and freely accessible provided in the Earth observation program Copernicus. Sentinel-2 satellite images covering 13 spectral bands. 10 classes with in total 27,000 labeled and geo-referenced images. The paper proposed network that achieved an overall classification accuracy of 98.57%. The geo-referenced dataset EuroSAT is made publicly available here [https://github.com/phelber/EuroSAT]. — 10 Classes - 27000 images 1. Annual Crop - 3000 images 2. Forest - 3000 images 3. Herbaceous Vegetation - 3000 images 4. Highway - 2500 images 5. Industrial Buildings - 2500 images 6. Pasture - 2000 images 7. Permanent Crop - 2500 images 8. Residential Buildings - 3000 images 9. River - 2500 images 10. Sea and Lake - 3000 images

Instantiate the EuroSAT dataset loader.

Parameters:
  • path – Dataset source path

  • variant'MS' (Multi Spectral) or 'RGB', Only MS inmplemented so far.

  • dtype – Data type for the input vectors