ISPRS, Vo!.34, Part 2W2, "Dynamic and Multi-Dimensional GIS”, Bangkok, May 23-25, 2001
several autonomous systems (Lockemann et al.,
1993). The main features of federated database
systems are heterogeneity and autonomy. Autonomy
means that a local databases management systems
of a federated database systems should not be
modified by the federated database systems and has
the right to decide which types of internal information
can be provided to the FDBS and to execute queries
and transactions according to its own rules ( Sheth,
A.P, 1990). A federated database can accept
different heterogeneous systems and integrate them
into higher level systems. They can be divided
between two basic forms of interaction (Lockemann
et al., 1993):
•Cooperation. Cooperation means several
systems or components use the same common data
source.
•Coordination. In case of coordination, the
desired data will be copied amongJhese systems or
components.
Fig. 2 Cooperation vs. Coordination
Interoperability refers in general to the ability of
various autonomous systems to bring together parts
and to operate in collaboration (ANDREJ
VCKOVSKI, 1998). In most cases this means the
exchange of meaningful information. Interoperability
has the following levels or types:
•Platform interoperability. It resolves the
differences in the hardware, system software, and
the services that deal with communication between
two objects ( Lockemann et al., 1997).
•Basic interoperability. It means that different
binary components that are developed by different
developers can interoperate each other.
•Versioning interoperability. When a
component is upgraded, all other system
components need not to be upgraded.
•Language interoperability. It is means that
components that are developed using different
languages can also communicate with each other.
•Semantic interoperability. It resolves the
communication between systems that are different in
the definition of categories, different in the definition
of the intension of classes and different in the
geometric description ( Bishr, 1997 ).
3. The heterogeneous Spatial Data
Spatial data include many kinds of
heterogeneous data. Vector, DEM and image will be
described here. Remote sensing is an important
source for vector data. It is able to cover large
regions and can provide temporally continuous
measurements. But most remote sensing systems
can not handle vectors and are unsuitable for
operations like network analysis.
On the other hand, vector data is the auxiliary
information of remote sensing image processing and
classification, which is used to automatically extract
semantic information. Thus the integration of remote
sensing data with vector data is very important for the
development of spatial information science.
Vector data structures represent spatial phenomenon
using continuous coordinate. The spatial
relationships among map entities are stored or are
computed when needed.
DEM is used to model, analyse and display
phenomena about topography or other surfaces. It
has two major data structures: rectangular grid (or
elevation matrix) and TIN (Triangulated Irregular
Network). Girds present a matrix structure which
records topological relation between data points.
Storage structure of digital computers (i.e. grid) can
be stored as a two-dimensional array of elevations.
TIN structures, on the other hand, are based on
triangular elements, with vertices at the sample
points.
Image data is stored as raster data structure. Raster
data structures tessellate space and assign each
spatial element a unique value.
Fig. 3 The Integration of VectorDDEM and Orthoimage