|
With the rapid development of spatial information technology, geospatial information data increase extraordinarily, which is distributed and heterogeneous, thus making the sharing and interoperability more difficult. As specifications published by OGC tend to be integrated, the geospatial information service is becoming a key technology in solving the data sharing problem. This thesis adopts service-oriented theory and technologies. Firstly, we use grid to integrate distributed heterogeneous geospatial databases. Secondly, we study the geospatial data distribution technology, and make the geospatial data services conform to OGC specifications. Finally, a unified operation interface is provided, through which geospatial data is able to be shared and interoperated.Grid could integrate distributed data effectively. Geospatial database’s supports to grid technology make more geographical information application could access grid services. This thesis analyzes the structure and key technologies of open source grid middleware, and then makes it support geospatial data. Firstly, in order to shield the heterogeneity of different geospatial database, we add the semantic description to geospatial data. Secondly, distributed geospatial data sources are integrated through virtual mapping, and a number of spatial operation functions are supported. Finally, according to the characteristics of geospatial data, Hilbert curve is used to optimize the query process.OWS released by OGC, is a geospatial service framework based on open standards. In this framework, geospatial data could be discovered, accessed, analyzed, and utilized based on Web. We study the geospatial data distribution technology, and analysis the implementation of OGC Web server, especially the extension mechanism of data source, including data storage pattern and the semantic description to spatial query. An OGC service distribution middleware is designed for BeyonDB, which is a self-developed spatial database. Experiment results show that system is more efficient with cache.Finally, a grid-based distributed heterogeneous spatial database management system is designed, which has functions including resource registration, spatial query, spatial data rendering, spatial data export and so on. Experimental results show that the integrated query is efficient. Geospatial data in BeyonDB is distributed through OGC Web Service, and can be accessed.
|