摘要：The aim of the paper is to present automated methods of discrimination of urban development forms using object-oriented classification in high-resolution images taken by the Ikonos satellite. The object-oriented classification makes possible to describe individual classes using not only the spectral reflection values but also the shapes, textures and topology of objects. The classification process as such is based on the theory of fuzzy sets. The research covered an area of 25 km, situated in central Warsaw. As a result of object-oriented classification, five classes of development typical of large cities were distinguished and described.
关键词：object-oriented classification ; land cover ; land use ; urbanized areas ; Ikonos ; Warsaw ; eCognition ; segmentation ; classification
Loading...
