CloudSEN12

This package contains minimum code to run the CloudSEN12 models☁️ proposed in Aybar et al. 2022 and Aybar et al. 2024.

pip install cloudsen12_models

Run: ```python

Read S2 image from Google Earth Engine

bandsgee = ['B1', 'B2', 'B3', 'B4', 'B5', 'B6', 'B7', 'B8', 'B8A', 'B9', 'B10', 'B11', 'B12'] imglocal = eeimage.exportimage('COPERNICUS/S2HARMONIZED/20240417T06463120240417T070110T40RCN', crs='EPSG:32640',transform=Affine(*[10, 0, 300000, 0, -10, 2800020]), bandsgee=bands_gee, geometry=box(55.325, 25.225, 55.415, 25.28))

Load model

modelv2 = cloudsen12.loadmodelbyname(name="UNetMobV2V2", weightsfolder="cloudsen12models")

Compute cloud mask

cloudmaskv2 = modelv2.predict(imglocal/10000)

Plot

fig, ax = plt.subplots(1,2,figsize=(14,5),sharey=True, tightlayout=True) rgb = (imglocal.isel({"band": [bandsgee.index(b) for b in ["B4","B3","B2"]]}) / 4500.).clip(0,1) plot.show(rgb,ax=ax[0]) cloudsen12.plotcloudSEN12mask(cloudmaskv2,ax=ax[1]) ``` <img src="https://raw.githubusercontent.com/IPL-UV/cloudsen12models/main/notebooks/exampleflooddubai_2024.png">

Full example at: runingee_image.ipynb.

models

With this package, the following models can be loaded: (with the function cloudsen12.load_model_by_name())

• cloudsen12 Model trained on the 13 bands of Sentinel-2 L1C in the CloudSEN12 dataset.
• cloudsen12l2a Model trained on the 12 bands of Sentinel-2 L2A in the CloudSEN12 dataset.
• dtacs4bands Model trained on the NIR, RED, GREEN and BLUE bands of Sentinel-2 L1C in the CloudSEN12 dataset. Bands: ["B08", "B04", "B03", "B02"]
• landsat30 Model trained on the common bands of Sentinel-2 L1C and Landsat 8 and 9 in the CloudSEN12 dataset. Bands: ['B01', 'B02', 'B03', 'B04', 'B08', 'B10', 'B11', 'B12'].
• UNetMobV2_V1 Model trained on the 13 bands of Sentinel-2 L1C in the CloudSEN12 dataset. The cloud masks of this model are included in CloudSEN12+ dataset.
• UNetMobV2_V2 Model trained on the 13 bands of Sentinel-2 L1C in the CloudSEN12+.

For examples about dataset use and training see cloudsen12.github.io.

Citation

If you find this code useful please cite:

``` @article{aybarcloudsen122024, title = {{CloudSEN12}+: {The} largest dataset of expert-labeled pixels for cloud and cloud shadow detection in {Sentinel}-2}, issn = {2352-3409}, url = {https://www.sciencedirect.com/science/article/pii/S2352340924008163}, doi = {10.1016/j.dib.2024.110852}, journal = {Data in Brief}, author = {Aybar, Cesar and Bautista, Lesly and Montero, David and Contreras, Julio and Ayala, Daryl and Prudencio, Fernando and Loja, Jhomira and Ysuhuaylas, Luis and Herrera, Fernando and Gonzales, Karen and Valladares, Jeanett and Flores, Lucy A. and Mamani, Evelin and Quiñonez, Maria and Fajardo, Rai and Espinoza, Wendy and Limas, Antonio and Yali, Roy and Alcántara, Alejandro and Leyva, Martin and Loayza-Muro, Raúl and Willems, Bram and Mateo-García, Gonzalo and Gómez-Chova, Luis}, month = aug, year = {2024}, pages = {110852}, }

@article{aybarcloudsen122022, title = {{CloudSEN12}, a global dataset for semantic understanding of cloud and cloud shadow in {Sentinel}-2}, volume = {9}, issn = {2052-4463}, url = {https://www.nature.com/articles/s41597-022-01878-2}, doi = {10.1038/s41597-022-01878-2}, number = {1}, urldate = {2023-01-02}, journal = {Scientific Data}, author = {Aybar, Cesar and Ysuhuaylas, Luis and Loja, Jhomira and Gonzales, Karen and Herrera, Fernando and Bautista, Lesly and Yali, Roy and Flores, Angie and Diaz, Lissette and Cuenca, Nicole and Espinoza, Wendy and Prudencio, Fernando and Llactayo, Valeria and Montero, David and Sudmanns, Martin and Tiede, Dirk and Mateo-García, Gonzalo and Gómez-Chova, Luis}, month = dec, year = {2022}, pages = {782}, }

```

Acknowledgments

This research has been funded by the DEEPCLOUD project (PID2019-109026RB-I00) funded by the Spanish Ministry of Science and Innovation (MCIN/AEI/10.13039/501100011033) and the European Union (NextGenerationEU).