A Conceptual Image Data Model For Image Information Retrieval

Abstract

Information modeling in image information systems and especially in geographical information systems must enable the end-users to under-stand the logical structure of the data stored in the integrated spatial databases. To do this a conceptual image data model is needed. Such a conceptual model must mirror the large data volumes in the spatial database and many interrelationships between the stored entities. Many approaches can be thought of. Here an object-oriented and relationship-free conceptual data model is chosen, in which inference rules are introduced to compensate the lack of stored object relations in the spatial database. Inference systems are normally slow, and in geographical information systems this is due not only to large data volumes but also to the fact that the geographical distribution of the objects has to be considered. To speed up the inference process objects extracted from the map are organized as lists where the objects are ordered in such a way that the objects which are geographically close to each other are close to each other in the ordered list.