The Key Technique Research Of Semantic-based RS Image Data Searching

With the rapid development of spatial information science, remote sensing technology has been developed into the spatial information technology including remote sensing, geographic information systems, global positioning systems and other technology. Means of access to a variety of spatial data showing the trends of diversification and speediness, In particular the growth of remote sensing image data is convenient for the access to space information, at the same time, the urgent problems that the huge amount of data for its image data management and application should to be solved. It is because of this, the research and discussion of massive image data management and application has become a hot topic in today’s remote sensing technology applications.In the analysis of the status of research on semantic-based remote sensing image data retrieval technology, for the storage, management, retrieval and distribution of remote sensing image data, the solutions are designed for semantic-based remote sensing image data retrieval.Primary work in this thesis can be summarized as following:1. Aiming at multi-source remote sensing image data, the metadata reference standard for semantics-based remote sensing data searching is proposed in this paper. The reference standard for the storage and management of remote sensing image data acquired has a certain universality, conducive to the retrieval and dissemination of image data, so as to promote the exchange and sharing of remote sensing data and improve the utilization of the image data.2. according to the characteristics of remote sensing image, metadata storage management is realized by metadata extraction technology, the storage for remote sensing image metadata and image data organization has conducted in-depth research, and given better solution for remote sensing image data storage management.3. Corresponding with the reference metadata standard for semantic-based remote sensing data searching, this paper constructs the model of remote sensing application ontology, defines concepts and attributes, and describes the relationship between them. Semantic query expansion of inclusion relations and equivalence relation is realized by means of querying remote sensing image searching application ontology.4. Integration of these results, semantic-based remote sensing image data retrieval prototype system called RsImageSearch is designed and implemented, elaborated on the process of the design and implementation of the various modules,to realize the automatic reconstruction and synchronization of remote sensing database index. The experimental results show that the prototype system achieved the fast retrieval of remote sensing image data and humanized WEB services based on natural language. A better solution is given to the problem of storage management, sharing, publishing and other issues of the mass remote sensing image data, and improves the results of search engine.