Full text: Proceedings, XXth congress (Part 5)

     
  
    
    
     
    
     
  
   
     
   
    
     
    
     
     
     
     
   
   
     
     
   
     
     
      
  
      
     
    
   
    
  
  
   
  
  
mbul 2004 
  
  
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part BS. Istanbul 2004 
similar view angle, and so on. Compare with cach other of the Point AX(m) AY(m) AZ(m) 
Figure pairs, one can access the effectiveness of reconstruction cere eer e ER 
of the 3-D model. 1001  -0.026 -0.060 -0.006 
1002  -0.037 -0.102 -0.041 
1003  -0.011 -0.003 -0.009 
1004 0.109 0.004 -0.087 
For verifying the geometric characteristic, the 3-D model has x 
also been exported and saved in a 3-D DXF file. The DXF file jos 105 OU 0065 
is then imported to the AutoCAD software, and a DWG file is 1006 | -0.014 -0.056 0.021 
drawn as shown in Figure 9. 1007 -0012 -0018 0.018 
+ Check point /| ee 1008 0.001 -0.010 0.014 
1009 -0.010 -0.035 0.007 
1010 -0.007 -0.020 -0.001 
1011  -0.004 -0.018 0.000 
1012  -0.0011 -0.031 -0.008 
1013 0.016 0.003 -0.007 
1014  -0.033 -0.049 0.032 
1015 0.041 0.102 0.009 
4. TESTING WITH GROUND TRUE 
  
Figure 9. Reconstructed CAD Model 1016 -0.046 0.057 0.046 
: 3 N : 1017 0.003 -0.042 -0.009 
3-D coordinates of the check points in Figure 9 can be queried 
using AutoCAD. As stated above, 3-D ground coordinates of 1018 -0.014 -0.041 -0.036 
the check points were also surveyed using a total station 1019 0038 0677 0007 
instrument. The differences of the two sets of coordinates are 
shown in Table 1. As Table 1 demonstrates, the RMSE of AX, 1020 0.073. 0.230 -0.009 
AY, and AZ coordinates are +0.033m, +0.070m, and +0.029m 1021 -6.051 0.191 -0.034 
respectively. 
1022  -0.007 -0.030 -0.028 
1023 0.014 0.010 -0.009 
5. CONCLUSIONS 1024 0.027 0.040 0.005 
: ; ; 1025  -0.002 0.028 0.025 
Pictures and ground control are two of the most important items 
in generating and reconstructing a 3-D model. When taking 1026  -0.021 -0.026 0.000 
pictures, the convergent angle between the exposure stations 1027 -0.018 -0054 -0010 
should be close to a right angle. However, if the pictures are 
used for texture mapping, the camera axis should keep nearly 1028 0.029 0.008 -0.001 
perpendicular to object surface. 1029 0.013 0.039 0.021 
When generating the 3-D model, reconstruction of curved edges 1030 -0.008 -0.092 0.046 
and surfaces is not an easy work. Take a vertical picture and a 1031 -0.017 -0.040 -0.009 
parallel picture for a curved edge and surface is recommended. EI : 
RMSE 0.033 0.070 0.029 
Each picture should cover as many control points as possible. Table 1 Coordinates Differences of 
In this study, although 34 points have been surveyed in field, 3-D Model and Field Surveying 
but only 3 of them appear on the most pictures. Therefore, the 
three points are selected for ground control points, the other 31 
points are used for check points. REFERENCES 
As shown in Table 1, the RMSE of Y coordinates is near twice Braun, C., 1999, Interactive Modelling Tools for 3-D Building 
ofthose of X and Z coordinates. This is reasonable for the Reconstruction, Photogrammetric Week ’99, Stuggart, Vol. 47, 
terrestrial and close range images taken in this study. pp.23-34. 
Eos Systems Inc., 2000. PhotoModeler Pro 4.0 User Manual. 
Gülch, E., 2000, Virtual Cities from Digital Imagery, 
Photogrammetric Record, 16(96), pp.893-903. 
http://ca.com/cosmo/ 
Shufelt, J. A., 1996, Exploiting Photogrammetric Methods for 
Building Extraction in Aerial Images. International Archives of 
  
	        
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.