hulaba / crop_mapping

Time-Independent Crop Type Mapping

Development of a transferable machine learning model for crop type mapping (in Europe)

(Time-Independent) Crop Type Mapping

Tasks

1. Get Sentinel 2 tile
2. Get it into R. (raster)
3. Calculate NDVI (RStoolbox)
4. Crop Points to Tile
5. Extract Sentinel 2 values at point locations (raster::extract)
6. Train Random Forest Model (caret, ranger)
7. Predict on Sentinel 2 tile

For code references: https://github.com/LOEK-RS/globalAOA

Geocomputation with R: https://geocompr.robinlovelace.net/spatial-class.html