Machine Learning-Based Processing Proof-of-Concept Pipeline for Semi-Automatic Sentinel-2 Imagery Download, Cloudiness Filtering, Classifications, and Updates of Open Land Use/Land Cover Datasets

Land use and land cover are continuously changing in today’s world. Both domains, therefore, have to rely on updates of external information sources from which the relevant use/land (classification) is extracted. Satellite images frequent candidates due their temporal spatial resolution. On contrary, extraction demanding terms knowledge base time. The presented approach offers a proof-of-concept machine-learning pipeline that takes care entire complex process following manner. Sentinel-2 obtained through pipeline. Later, cloud masking performed, including linear interpolation merged-feature time frames. Subsequently, four-dimensional arrays created with all potential training data become basis for estimators scikit-learn library; LightGBM estimator then used. Finally, classified content applied open databases. verification provided experiment was conducted against detailed cadastral data, Shannon’s entropy since number cadaster classes naturally consistent. showed good overall accuracy (OA) 85.9%. It yielded map study area consisting 7188 km2 southern part South Moravian Region Czech Republic. developed replicable any other interest so far as requirements input met.