SEN12TS: Sentinel SAR and Optical Land Cover Image Triplets for 2020
Updated 3y ago
Available on 1 platform
Sign in to view source links and access this dataset
Description
SEN12TS is a 1.69 TB dataset from MLHUB containing 246,400 triplets of Sentinel-1 SAR, Sentinel-2 optical, and land cover imagery. It covers six agro-ecologically diverse regions—California, Iowa, Catalonia, Ethiopia, Uganda, and Sumatra—across 31,398 unique spatial tiles at 10m resolution. The dataset was last updated on January 1, 2023.
Use Cases
Training generative adversarial networks to transform radar imagery into enhanced vegetation indices, as mentioned in the documentation.
Testing combinations of SAR and optical input imagery for cropland classification tasks.
Developing time-series models for land cover change using radar imagery collected at four 12-day intervals.
Benchmarking geospatial machine learning models on a large-scale, multi-region dataset.
Analyzing the relationship between SAR backscatter, InSAR coherence, and multispectral optical data for specific land types.