Full text: Proceedings, XXth congress (Part 4)

nbul 2004 
art of the 
‘II data of 
solution). 
raw data. 
using the 
arises the 
tes of the 
:d for the 
  
JEM data 
ion DEM 
next part 
» 
Jo 
we need a 
base. This 
isses and 
rocessing 
as newly 
accessing 
rvice for 
requested 
egion the 
T in the 
f the final 
le to load 
ution into 
rh enough 
ırge DEM 
rformance 
hal scene. 
s can be 
time. 
  
ce decides 
umber of 
DEM and 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B4. Istanbul 2004 
  
looks for available data in the database, beginning with the 
highest possible resolution. One aspect we have to consider 
is, that the database has no or only fragmented data for the 
requested region. Possible reasons are: 
e The acquisition of the laser scanner data is still 
going on. That means the requested data is not yet 
fully available. 
e The processing of the laser scanner data is not yet 
fully completed. 
In this case the Height-Service will search for a lower 
resolution digital elevation model. Now we are ready for the 
access of the DEM data. 
2.6.2 Accessing the DEM data 
The Height-Service includes methods for accessing the data 
from the selected DEM. The class diagram shows the most 
important methods. 
TERRE] 
1 
<<Interface>> | 
HelghtProvider TELE IOINCITENETISIMUIUEUSTUTUMEM 
3 rod BoundingBox | InvalidBoundingBoxException 
1 200.” 
HelghtOperations 
    
boundingBox 
terWidthNorth{) 
os 
: ««Interface»» 
|. HeightinterpolateFunction 
| 
"NearasiNeighborinterpolateFuncfion | 
[ LinearinterpolateFunction | 
  
Fig. 2: Height-Service class diagram 
The methods: 
® getCoordinates (double [] 
eastValues, double[] northValues) 
e getHeightValues () 
are accessing the grid structured base points of the DEM, 
while the methods: 
e getNearestNeighbourInterpolatedHei 
ght (double east, double north) 
e getLinearInterpolatedHeight (double 
east, double north) 
are able to interpolate height values between the base points. 
2.7. Summary 
Reviewing the data handling section of this paper we began 
with a look at the data basis. Then we created PNG-image 
files from the raw data and stored them in the database using 
the raster tiling approach. Finally we considered the access to 
the data in the database via the Height-Service interface. That 
means we are now ready for the next part of the paper: the 
visualisation. 
519 
3. VISUALISATION OF THE DATA 
This part of the paper concentrates on the visualisation of the 
DEM data. The presented concepts of this part are not limited 
to the visualisation of our Baden-Württemberg example. The 
techniques can be used for all kinds of rasterised terrain 
visualisations. 
Most 3D applications concerning environmental aspects need 
digital elevation models as a data basis. The visualisation of 
these models leads to a three-dimensional terrain-object 
presentation in a virtual reality. 
3.1 Creation of the Virtual Universe 
For the integration into the base GIS GISterm, the virtual 
universe is created with the help of the 3D-Service, which 
uses the means of Java 3D. This virtual universe includes 
common functions and properties needed for every 3D-GIS 
application: 
e Management of semantically grouped object 
content in different 3D layers. 
e Means for orientation and navigation. 
e Possibility for changing common scene properties, 
like the disbanding of the height axis, the change of 
the lighting or the background colour. 
For the visualisation of the digital elevation models we need 
a 3D Layer, holding the terrain object generated from the 
DEM data out of the database. This is described in the next 
section. 
3.2 Creating Three-Dimensional Object Content 
For visualising the DEM data in the virtual universe, a 3D 
Layer called DEMLayer3D is created. This Layer holds the 
actual 3D object content. The DEMLayer3D is added into 
the virtual scene created by the 3D-Service, which visualises 
the terrain in the 3D-viewer component. 
3.2.1 Structure of Terrain Object 
This part of the paper describes the structure of this terrain 
object. Due to texture size issues explained deeply in the next 
section the terrain objects need to be divided into several 
parts. The number of terrain parts depends on the size of the 
texture parts. By dividing the terrain size by the texture size 
the number of needed terrain parts is calculated. The size of 
the terrain parts corresponds with the size of the texture parts. 
The arrangement of the terrain parts starts in the lower left 
corner increasing into east and north direction. The terrain 
parts at the upper north and east border are smaller than the 
texture size. The following image shows the arrangement of 
the terrain parts. 
   
Fig. 2: Arrangement of terrain parts 
 
	        
Waiting...

Note to user

Dear user,

In response to current developments in the web technology used by the Goobi viewer, the software no longer supports your browser.

Please use one of the following browsers to display this page correctly.

Thank you.