Extraction of land cover from ZY-3 imagery using ontology-based method

In this paper, an ontology-based framework of China Geographic National Conditions Monitoring (CGNCM) land cover extraction is presented. Using statistical analysis method, spectral, texture, spatial features of each land cover class are extracted from the referenced ZY-3 image acquired in summer, 2012. Stored in the form of OWL, the features of one class can be considered as the prototype of this class. With the help of prototype, automation extraction of land cover information is validated on ZY-3 images of summer in subsequent year. The results highlight the capacity of the ontology-based method to automatically identify land cover using the ZY-3 image in CGNCM mission.