GEE LULC Toolkit
A single-window QGIS workflow for LULC mapping with Google Earth Engine: load composites and spectral indices, create stratified sample points via 4-view labeling (custom schemes or ESA WorldCover), apply classifiers (RF/GBT/CART/SVM) or unsupervised clustering, and output Olofsson area‑adjusted accuracy with 95% CI, Jeffries–Matusita separability, plus an automated map layout builder.
An end-to-end land-use/land-cover (LULC) classification workflow powered by Google Earth Engine inside QGIS. One combined window guides you from image loading to a finished map: composites and spectral indices, stratified sample points, point labeling (LULC, ESA WorldCover, or custom schemes) with a 4-view labeling window, Random Forest / GBT / CART / SVM or unsupervised clustering, Olofsson area-adjusted accuracy with 95 percent confidence intervals, Jeffries-Matusita separability, and automated map layout builder. Requires the earthengine-api Python package in the QGIS Python environment and a Google Cloud project registered for Earth Engine.