International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part Bl. Istanbul 2004 
  
These procedures and products are represented by the following 
software components: 
TLS-SMS: User interface; image measurement in mono and 
stereo; 3-ray forward intersection (point positioning); image and 
shadow enhancement 
TLS-IRS: Quasi-epipolar rectification to plane or via 
DSM/DTM and ortho-image generation 
TLS-LAB: Sensor/trajectory modeling, triangulation; auto- 
matic and semi-automatic tie-point generation 
TLS-IMS: Image matching for DSM/DTM generation, 
DSM/DTM modeling and interpolation 
Adaptation of methods and software for ALOS/PRISM 
processing 
Feature/object extraction, e.g. city modeling: CC-TLSAutotext. 
A 3D city model obtained through 3D object extraction and 
modelling can be draped with textures as shown in Figure 10 
(Gruen et al., 2003). 
  
  
Figure 10. A sample 3D model of Shin-Yokohama 
3.2 Image GIS 
The consequent integration of images and their processing 
potential into GIS (Geographic Information System) has not 
been successfully achieved yet. We present here a concept 
called “Image GIS", whose emphasis is on combining image 
measurement functions with GIS database functionality on one 
unique platform. Figure 11 shows an example of such a Image 
GIS viewer. An ortho-image may serve as a background for a 
2D map that is obtained from another source. The system 
allows one to make measurements, like distances and areas on 
both the map and the image. In addition, the Image GIS viewer 
is equipped with a mono-image measurement system, where, as 
the center of the ortho-image moves, the target images for the 
3D measurement system move as well. When one wants to 
measure the 3D position of a point precisely by pointing to it on 
the ortho-image, the viewer brings one with its corresponding 
image set (forward, nadir and backward) to the approximate 
position for the 3D measurement. The result can be stored in a 
database management system (GeoBase with MSDE and 
Access). The Image GIS viewer can for example measure the 
cross-section of a line-shaped object like a river, road, railway, 
etc. by combining automatic sampling and semi-automated 
matching as shown in Figure 11. 
     
3D Viewer 
2D Viewer 
3 T ES Cross-section measurement i 
Cross-section Viewer Ba a RS 
Figure 11. Image GIS Viewer 
KA mim 
4. APPLICATIONS 
The new features of the TLS system and the image data are 
stimulating more demands for 3D seamless information 
extraction of line-shaped objects and landscape modelling at 
high resolution. Line-shaped man-made objects include roads, 
bridges, railways, power cables, pipelines, etc. Images and 
spatial information are needed for investigation before the 
construction of those, maintenance and management after the 
construction, and as base data for a variety of GIS applications 
(Tsuno, 2002b). Figure 12 is an example for the investigation of 
the environment of a river, and can be used for the investigation 
of vegetation, the gravel grain diameter distribution of a dry 
riverbed, the river floor profile, etc. (Fukami et al., 2002). 
    
     
in Yoni-cho, Saitama 
Prefecture 
e * A € > * 
Med N 
BR RS 
on a contracted research from. i = + : ue 
ependent. Adminktratze Instibtiore S “2 
Public Works Fe search Instinte) 
      
Figure 12. River environment investigation 
TLS imagery can also be applied to generate base ground data 
for flooding simulation with high precision, in order to judge 
whether water goes over a Japanese raised floor or below. High 
definition 3D city models are used in city planning, landscape 
simulation, auto-navigation, gaming and so on. When the 
helicopter flies along a road, wall texture facing the road can be 
acquired with a nadir-looking or oblique-looking image. The 
wall textures that are perpendicular to the road can be acquired 
with either forward-looking or backward-looking images. 
Textures can be semi-automatically mapped onto 2D polygons 
in 3D space. 
Figure 13 is an example of a true ortho-image (TrueOrtho 
where the TLS data is taken with a high overlap between 
neighbouring strips. The data has been processed in cooperation 
with ISTAR, France. 
Using the characteristics that images of the same area can be 
taken with a constant time lag, the system can measure the 
i 
    
  
    
  
  
   
    
   
     
    
    
   
     
   
    
  
  
  
  
  
  
   
   
   
  
   
    
  
    
   
   
  
  
  
  
  
  
  
  
   
    
     
    
      
Intel 
—— 
spee 
flow 
prob 
Sem 
as b 
com 
imp 
desc 
traci 
assis 
to be 
  
  
^A 7 
pos 
bet 
the 
mai 
of 1