International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part BS. Istanbul 2004 
3. THE PRODUCTION OF 3D MODEL BY HIGH AND 
LOW AERIAL PHOTO AND BY GROUND PHOTO 
This is how we actually produced the 3D model. The object is 
the Topcon main building. 
The pictures we used were: the aerial photo by the film camera 
of airplane, the photo by digital camera of powered paraglider 
and the lateral photo obtained by VGS on the ground. We first 
produced 3D model for each of them by PI-3000 and then fused 
them together. As for the control points, we first measured the 
side of the building by non-prism TS and converted the data to 
GPS coordinate using the points measured by GPS in the 
building lot. Table 4 shows the capacity of the cameras used for 
high, low and ground level photographs. Table 5 shows the 
conditions for the analysis by each camera. And Figure 5 is the 
display of PI-3000 showing all of the analysis of each camera. 
Here is how each photograph and analysis was processed. 
  
   
  
  
   
  
  
  
      
  
  
  
  
  
   
  
   
  
Airplane Paraglider VGS 
Camera WILD * CANON*°° Minolta * 
RC-20 EOS* * Digital Dimage7 
Sensor Film C-MOS(22.7 CCD 
* 45.Imm) * 9/3inch* * 
Number of | 6062 x 5669 3072 x 2048 2568 x 1928 
Pixels (by Scanner) 
Resolution 42.32: m 7.4. m 3.4 m 
Focal 152.4mm 18mm *» 7. 4mm 
  
  
  
length 
  
  
   
  
    
Table 4. Cameras used for high, law and ground level 
  
  
  
   
  
  
  
  
   
  
  
  
  
  
  
Airplane Paraglider VGS 
Photographing 2000 x 2000 200 x 100 30x25 
Area 
Modelling Area 300 x 300 100 x 50 20 x 20 
Altitude 1609 163 26(side) 
Base length 886 18 5 
Resolution: AXY 0.89 0.06.6 0.01 
Resolution: AZ 1.61 0.6 0.06 
  
  
  
  
  
     
   
  
   
Table 5. Condition (unit: m) 
  
# 
    
  
       
   
    
  
  
: 
ë ___Aenal Phocographine Fest. 
2 
| JE (Fiswered Pata-tlsder i 
°° : Pu. Muxlelug Area 
i LAbou 300 x 300m) 
FE bol n 
s ; 1* > p Aerial Photographing Ares 
"s X (Powered Pur ice} 
  
  
Figure 5. All of the analysis of each camera 
         
   
  
  
   
  
    
  
3.1 Making of the base by the analysis of high altitude 
aerial photo 
For aerial photo, we used the photo which shows 4 kilo-meter 
square around Topcon in Itabashi ward of Tokyo. The camera 
was WILD ** RC-20** with focal distance of 150mm. The 
altitude was 1609m. Out of the base length (distance between 
cameras) 886m we picked out the area 300m* 300m for 
modelling (Figure 6). Since we used the photo of the high 
altitude, which was taken by film, we got its analyzed data by a 
scanner with 600dpi and put them into PI-3000 for inner 
orientation. The Figure 7 is the modelled image processed by 
PI-3000 to be used as the base of all operation. This time we did 
not make stereo-matching but made simple 3D modelling. 
However, if there is such thing as mountain, we have to make 
stereo-matching on that area and produce a 3D model. 
  
  
T LP ue a a Ra i 
RU reb a ur RA He as 
ui guet PAS AP Te ci a 
ET EE i : 8 E Ord a A 5 
5 Mur: 
fa 
^ #3 
Sp 
TE Me op 
ern 
; 
  
    
Figure 7. Modelled Aerial Image (300m x 300m) 
  
  
   
Internation 
3.2 The] 
lower alti 
This time 
parachute 
digital can 
that it can 
not legally 
The came 
sensor of 
15.1mm, | 
length 181 
100mx50r 
The analys 
We made 
measured | 
foregrounc 
automatic 
plane (Fig 
finally we 
10(d)).