Skip to content

CloudSEN12 Global dataset for semantic understanding of cloud and cloud shadow in Sentinel-2

CloudSEN12 is a large dataset for cloud semantic understanding that consists of 9880 regions of interest (ROIs) that consists of 49,400 image patches (IP) that are evenly spread throughout all continents except Antarctica. Each IP covers 5090 x 5090 meters and contains data from Sentinel-2 levels 1C and 2A, hand-crafted annotations of thick and thin clouds and cloud shadows, Sentinel-1 Synthetic Aperture Radar (SAR), digital elevation model, surface water occurrence, land cover classes, and cloud mask results from six cutting-edge cloud detection algorithms. Each ROI has five 5090x5090 meters image patches (IPs) collected on different dates that match one of the following cloud cover groups:

  • clear (0%)

  • low-cloudy (1% - 25%)

  • almost clear (25% - 45%)

  • mid-cloudy (45% - 65%)

  • cloudy (65% >)

The dataset is available here. For more details check out the website and you can read the preprint of the paper here

Data Citation

Aybar, C. et al. CloudSEN12 - a global dataset for semantic understanding of cloud and cloud shadow in Sentinel-2.
Science Data Bank https://doi.org/10.57760/sciencedb.06669 (2022).

Paper Citation

Aybar, C., Ysuhuaylas, L., Loja, J. et al. CloudSEN12, a global dataset for semantic understanding of cloud and cloud shadow in Sentinel-2.
Sci Data 9, 782 (2022). https://doi.org/10.1038/s41597-022-01878-2

Currently included layers are:

Earth Engine Snippet: Hand-crafted labels - high-quality

var cs12_high = ee.ImageCollection("projects/sat-io/open-datasets/cloudsen12/high");

Sample code: https://code.earthengine.google.com/?scriptPath=users/sat-io/awesome-gee-catalog-examples:global-landuse-landcover/CloudSEN12-HIGH-QUALITY

Earth Engine Snippet: Hand-crafted labels - scribble

var cs12_scribble = ee.ImageCollection("projects/sat-io/open-datasets/cloudsen12/scribble");

Sample code: https://code.earthengine.google.com/?scriptPath=users/sat-io/awesome-gee-catalog-examples:global-landuse-landcover/CloudSEN12-SCRIBBLE-QUALITY

Earth Engine Snippet: Hand-crafted labels - nolabel

var cs12_nolabel = ee.ImageCollection("projects/sat-io/open-datasets/cloudsen12/nolabel");

Sample code: https://code.earthengine.google.com/?scriptPath=users/sat-io/awesome-gee-catalog-examples:global-landuse-landcover/CloudSEN12-NO-LABEL

Earth Engine Snippet: IPs footprint

var cs12_geom = ee.ImageCollection("projects/sat-io/open-datasets/cloudsen12/footprint");

Sample code: https://code.earthengine.google.com/?scriptPath=users/sat-io/awesome-gee-catalog-examples:global-landuse-landcover/CloudSEN12-FOOTPRINT

License

This work is licensed under a Creative Commons Attribution 4.0 International License. You are free to copy and redistribute the material in any medium or format, and to transform and build upon the material for any purpose, even commercially. You must give appropriate credit, provide a link to the license, and indicate if changes were made.

Curated in GEE by: Samapriya Roy

Keywords: cloud, deep learning, Sentinel-2, Sentinel-1, U-Net

Last updated: 2022-09-18