BetaBetaDataset
To enable accurate, scalable mapping of agricultural field boundaries across the world using satellite imagery and machine learning. It provides global-scale predictions and satellite-derived inputs for agricultural fields (2024–2025)
A global AI-generated map of farms and field boundaries, built from satellite images, with both raw imagery and final predictions included.
Satellite images, AI-generated field maps, vector polygons of farm boundaries, and web-ready map tiles.
Built using a combination of satellite remote sensing, deep learning (machine learning), and large-scale geospatial processing pipelines. Use satellite images, clean and prepare them, run a deep learning model to detect farm fields, and finally convert the results into global maps