Archived: 3 December 2001
Adelaide University Honours thesis, October 2001.
Supervisor: P.D. Coddington
The U.S. National Imagery and Mapping Agency (NIMA) has developed the Geospatial and Imagery Access Services (GIAS) specification to address this problem. Similar efforts are being made to develop interface standards by the OpenGIS Consortium (OGC).
Searching multiple archives is problematic as each may employ different image storage formats or metadata schemas (data models). This research shows how XML and XSLT, the emerging standards for data description and exchange, can be used to solve the problem of converting between different data models in a distributed geospatial image archive. This is demonstrated by the design and implementation of a new metadata conversion architecture for OLGAS, a Java program that conforms to a subset of the GIAS specification. The data models supported in OLGAS were defined using XML Schemas and the translations between different data models were achieved using XSLT. The metadata conversion architecture discussed in this paper enables federated querying across multiple image libraries that employ different data models. The architecture may also be used in libraries that support the Open GIS Consortium interface standards.