Využití radarových dat ze Sentinelu 1 pro klasifikaci LandCover

Abstract

Bachelor thesis focuses on the processing and classification of radar records. Used radar records were captured by a Sentinel 1A satellite that carries a C-SAR single-frequency radar sensor. Recordings were made in 2016 for the growing season from May to August. The thesis uses an object-oriented classification of agricultural land used for two areas of interest. The legend was based on the LPIS database, three classes were selected - arable land, permanent grassland and a class including vineyards and fruit orchards. The radar records were preprocessed an classified in the SNAP software, subsequent evaluation was performed in ArcMap and Excel software.