This repository is about crop mapping workflow for both numeric data using open datacube satellite data and image data extracted from xarray datasets.
- The workflow for the Numeric models and vision models is detailed in this repository.
- The workflow of mapping visualizations is detailed i the predictions to map notebook.
- The models are stored in pickle files in the model folder.
mapping_preview.mp4
-
Data extraction with STAC (accessing data in local or remote instance)
-
Feature layers modification (choosing the optimal features for annual, in season models and changing the feature layers script)
-
Large image(zone) predictions to map (handling large zones mapping)
-
Adding the streamlit app (The streamlit app code and preview is ready, it will be added very soon to this repository)
- Exploring drought indices and water scarcity indices extraction and forecasting