LEE RS Lab
A remote sensing teaching lab in QGIS (English / Khmer): free Sentinel-2, Landsat, Sentinel-1, MODIS, DEM and land-cover data without any account or login, and the whole workflow from study area to classification, accuracy, change detection and report.
LEE RS Lab is built for teaching and learning remote sensing step by step. Choose a study area from administrative boundaries (any country; Cambodia preset with Khmer names) and the plugin downloads cloud-masked composites of only the pixels inside it from Microsoft Planetary Computer and Earth Search: Sentinel-2, Landsat 5-9, Sentinel-1, MODIS NDVI, Copernicus DEM, ESA WorldCover and JRC surface water. One guided panel then covers band combinations, spectral indices, filters, PCA, K-means / ISODATA, supervised classification (Random Forest, SVM, kNN, Maximum Likelihood, Minimum Distance, SAM, Parallelepiped), accuracy assessment with Cochran sample size and area-adjusted estimates (Olofsson et al. 2014), from-to change detection with a Sankey diagram, map layout and a PDF report. The LEE tab (Local Earth Engine) runs Earth-Engine-style scripts on your own computer, with 32 examples. A Basic / Advanced switch keeps the panel simple for beginners, and a magic wand creates training polygons with one click. Every section has a short lesson, and the whole interface switches between English and Khmer. A 15-lesson Khmer manual with annotated screenshots is at https://khgeo.github.io/lee-rs-lab/ . Optional: Random Forest, SVM and kNN need scikit-learn, which the user can install from 1 Data > Setup (pip, only when the button is pressed).