National high-resolution cropland classification of Japan with agricultural census information and multi-temporal multi-modality datasets

Junshi Xia, Naoto Yokoya, Bruno Adriano, Keiichiro Kanemoto

Abstract: Highlights•Cropland mapping utilizing national agricultural census data was proposed.•The first work used multi-temporal Landsat and Sentinel data to map cropland in Japan.•Deep learning and machine learning methods for cropland mapping were compared.•Our product offers the highest resolution and SOTA cropland map at the parcel level.