Transcript spatial_db

Joseph Jaquish
1/10
- Spatial Data Types
- Spatial Database Architectures
- Geographic Information Systems
2/10
- Spatial Data Types (SDTs)…not Spatial Types of
Data (STDs)
- Compared to Image Data Types
- Values in SDTs contain Spatial Information and
Non-Spatial Information (dimensions of building
and name of building)
- Road data type contains intersections, speed
limits and stop signs
- This is why mapquest.com is so smart!
3/10
- Single Objects
-Point: Location, not magnitude
-Line: Connection or Moving Through
-Region: Area/Extent in 2D space
- Spatially Related Collections of Objects
-Partition: Collection of Regions
-Network: Graph, Points=Nodes Lines=Edges
4/10
- Early architectures extended existing
database systems in two ways:
- Layered Architecture
-Dual Architecture
Spatial Tools
(Construct SDT)
Standard DBMS
Integration Layer
Standard DBMS
(Info Part)
Spatial System
(Geometry Part)
5/10
- Extensible architecture
- Key Idea: Treat SDTs like everything else!
- Store SDT as an object
- Use operator overloading
- string1+string2
building1+building2
6/10
Selections
- Return objects from larger set that fulfill a
search criteria
- Find all cities in Bavaria
- cities select[city.center inside Bavaria]
- Assuming 1) “inside” is provide with SDT
- 2) Bavaria is a region
7/10
Joins
- Compares two objects with a predicate
according to their spatial values
- Combine cities with their states
- cities states join[city.center inside state.area]
-For each river, find all cities within 50 miles
- cities rivers join[dist(city.center, river.route)<50]
8/10
Indexing
- 2 ways to form a key for an SDT
- Degree of Error: Maximum allowable
difference between objects
-Bounding Box
- Grid Approximation
9/10
- Forms scenes user chooses
- Bottom Up vs. Top Down
- Bottom Up: Parts to Whole
- Top Down: Whole to Parts
- OASIS, Real Time
Weather/Flight Simulation…wow…
10/10