A high-performance deep feature encoding network using GeoFusion-ChangeNet for land cover mapping and change detection in remote sensing imagery
Remote sensing-based land cover classification and change detection are essential for ensuring environment, planning cities, and managing resources. Accurately extracting spatial and temporal information from high-resolution multi-temporal images remains challenging due to feature inconsistency, class imbalance, and...