| In recent years, with the growing development of remote sensing technology, the ability to get remote sensing data has also been improved, as a result, the remote sensing technology has provided plenty of data and convenience for all kinds of applications. Although the acquisition ability of remote sensing data has been improved, there are still weakness on full utilization of remote sensing data especially the spatial topological information. In order to solve this problem, on the basis of analyzing the classification of remote sensing image, a construction method of sample database by dividing the sampling region and a construction method of knowledge database by taking the geographic ontology as the model have been proposed, which solve the expression problem of spatial topological relations between the objects in sampling region effectively.The main research achievements include the following aspects:(1) A new remote sensing images sampling method by dividing sampling region has been proposed in this paper, and then construct the sample database based on it. The method could extract spatial relationships between samples, and it makes up for the absence of spatial topological relationships using previous sampling method.(2) A new construction method of remote sensing image knowledge database based on geographic ontology models has been proposed. In this paper, we introduced the ontology theory into the field of remote sensing image expression. The geographic ontology is better than other knowledge expression models both in the hierarchy and reasoning ability, which meets the application needs of this paper.(3) On the basis of construction principle of geographic ontology, we explored the construction process of geographic ontology using remote sensing sample database, and then analyzed and summarized the mapping relation between sample database and geographic ontology to reduce the manual construction time of geographic ontology significantly. |
PDF Full Text Request