- Step 1: Prepare an AOI file and modify the GT data (QGIS)
- Make a geojson file of an AOI in EPSG:4326.
- Clip and reproject (e.g. EPSG:28350 in Albany) the GT data and rasterize it.
- Load the GT data
Raster > Extraction > Clip by mask layer
Raster> Projections > Warp (reproject)
(the output resolution should be matched to that of S2)- Export as geotiff
- Step 2: Data downloading and preprocessing (
S2_download_and_preprocessing.ipynb
)- Download an ECARR and NDVI image via GEE
- Preprocess the GT data
- Preprocess the index data (reprojection, shape adjustment, etc)
- Step 3: Main analysis (
eucalyptus_classification.ipynb
)- Train and validate the model for eucalyptus classification (logistic regression)
- Export the result as a vector map
![workflow_eucalyptus_classification](https://private-user-images.githubusercontent.com/159540763/316308436-5c3bae42-924b-4375-8d80-e3832bc4a6eb.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3Mzk5MTkxMTAsIm5iZiI6MTczOTkxODgxMCwicGF0aCI6Ii8xNTk1NDA3NjMvMzE2MzA4NDM2LTVjM2JhZTQyLTkyNGItNDM3NS04ZDgwLWUzODMyYmM0YTZlYi5wbmc_WC1BbXotQWxnb3JpdGhtPUFXUzQtSE1BQy1TSEEyNTYmWC1BbXotQ3JlZGVudGlhbD1BS0lBVkNPRFlMU0E1M1BRSzRaQSUyRjIwMjUwMjE4JTJGdXMtZWFzdC0xJTJGczMlMkZhd3M0X3JlcXVlc3QmWC1BbXotRGF0ZT0yMDI1MDIxOFQyMjQ2NTBaJlgtQW16LUV4cGlyZXM9MzAwJlgtQW16LVNpZ25hdHVyZT01ZTVjMzk3NDllYmFkMWQ2YTc0MDRmNGUzZThjN2JkMjg5ZmVmYTU3ZmNhOGYyZTRlZTNiYjEwNWIyNzMzYzA0JlgtQW16LVNpZ25lZEhlYWRlcnM9aG9zdCJ9.lmaEVEUPxp7WRl-PAtS5yDNEQwKxYN7pzVId75xR2IA)