Transcript Document

Spatial Information Systems (SIS)
COMP 30110
Formats and Standards
Standards and Formats
• GIS vendors provide their own proprietary formats
Examples:
– Shapefile (ESRI ArcView format): based on a non-topological representation
(as for spaghetti data structure)
– Coverages (ESRI ArcInfo format): based on a topological representation
– DXF (Drawing Interchange Format: developed within Autodesk’s AutoCAD
sw)
– etc.
Shapefiles
• Shapefile (ESRI ArcView format): based on a non-topological representation
(as for spaghetti data structure)
• Topological/connectivity relations are calculated on-the-fly
• A map is composed of different layers (non-overlayed approach)
Idrography layer
Road network layer
Entities from the
two layers
intersect
Note
•ArcView is a desktop GIS:
– Used for visualisation more than for map making
– Does not provide all full functionality provided by ArcInfo
– Easily integrates with other applications, DBMSs etc.
• ArcInfo is a dedicated GIS system:
– Used more for map making
– Provides complete GIS functionality
– It was not designed to be integrated with other applications
– No high-level (SQL-like) query language
Standards and Formats (cont.d)
•Formats defined by official organisations (standards):
– TIGER (Topologically Integrated Geographic Encoding and Referencing):
system and DB developed by the US Census Bureau: based on topological
representation
– SDTS (Spatial Data Transfer Standard: developed by the US Geological
Survey): based on a complex topological representation
– etc.
TIGER files
• Based on topological representation
• All objects in one single layer: overlayed approach
• All intersections are stored explicitly even if they do not correspond to
geographic objects
This intersection
point is stored
explicitly (i.e.,
lines are split)
TIGER files (cont.d)
• Based on topological representation
• Entities:
– points
– chains (endnodes and shape points)
– polygons
• Relations:
– VE
– FE
– EV
– EF
Additional information: polygon centroids, attribute information,
isolated points, dangling edges, etc.
Other ways of representing vector data:
Half-plane representation
• Half plane representation: polygons are defined as intersection of
a number of half planes (each corresponding to one of their sides)
• The points that belong to the interior of the polygon satisfy the constraints
aix+biy<0 (corresponding to half planes)
Example:
x-y>0
^
x<7
^
y>1
^
y<3
^
Other ways of representing vector data:
Realms (Güting and Schneider 1993)
• Realms: planar graphs defined over discrete domains (i.e., grids - not
the Euclidean plane)
• Realm objects: points, lines and regions defined in terms of finite
representations
• Lines and regions defined in terms of realm points and segments
• Intersections of lines occur only at realm points
Realms (cont.d)
• Intersections of lines occur only at realm points
• Therefore a realm is represented by means of a set of points and a set
of non-intersecting segments (they only “touch” at their endpoints)
Realms (cont.d)
• A realm
Realms: remarks
• More complicated sets of data: not just a polygonal subdivision but
also isolated points and dangling edges (inside and outside faces)
• Classical data structures (e.g., DCEL) must be extended to be able to
capture these cases