International Archives of the Photogrammetry, 
Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004 
Two options to obtain the epipolar geometry of the stereo model 
have been shown. While the first one requires two homography 
matrices for the recovery of the fundamental matrix and the last 
require only one, but provide two different solutions. Both 
options lead to the need of at least two planar curves to get 
unique solution. Equation (9) provides 6 linear equations for 
each homography matrix so Least squares adjustment has to be 
performed having two or more homography matrices. Equations 
(14) provide two solutions for R and T for each homography 
matrix. When having more than one homography matrix we 
need some elimination procedures to get the right solution. 
4. SUMMARY 
Two methods for recovering the relative orientation parameters 
from free-form planar curves have been presented and tested. 
While both perform quite well in most cases, the homography- 
based algorithm exhibited more robust behavior in terms of not 
being sensitive to some singular configurations observed for the 
alternative method. A forthcoming paper on the subject will 
present a more thorough analysis. of the aforementioned 
methods including experiments with real data. 
REFERENCES 
Besl, P. J., and McKay, A Method for registration of 3D shapes. 
IEEE transaction on pattern analysis and machine intelligence. 
Vol. 14, no. 2, Feb. 1 
992. 
Fitzgibbon, A.W., Robust registration of 2D and 3D point sets. 
In proc. British machine vision conference. Manchester, sep. 
2001. pp. 411-420. 
Habib, A., and Kelley D., Single-Photo Resection Using the 
Modified Hough Transform. Photogrammetry Engineering and 
Remote Sensing. Vol. 67, No. 8 August 2001, pp 909-914. 
Habib, A., and Kelley, D., Automatic Relative Orientation of 
Large Scale Imagery Over Urban Areas Using Modified Iterated 
Hough Transform. 2001. /SPRS Journal of Photogrammetry 
and Remote Sensing 
Vol. 56, pp. 29-41. 
Habib, A., and Lee, Y., and Morgan, M., Automatic Matching 
and Three-Dimensional Reconstruction of Free-Form Linear 
Features from Stereo Images. Photogrammetry Engineering and 
Remote Sensing. Vol. 69, No. 2, February 2003, pp. 189-197. 
Hartley, R., and Zisserman, A., Multiple view geometry in 
computer vision. Cambridge university press. 2000. 
Kraus, K., Photogrammetry vol. 2, Ferd. Dummler Verlag. 
Bonn 1997. 
Kubik, K., Relative and Absolute Orientation Based on Linear 
Features. 1988. ISPRS Journal of Photogrammetry and Remote 
Sensing Vol. 46, pp 
. 199-204. 
Ma, S.D., Conics-Based Stereo, Motion Estimation, and Pose 
Determination. 1993. International Journal of computer vision 
10:1 pp. 7-23. 
Mikhail, E., Linear features for photogrammetric restitution and 
Integrating Photogrammetric Techniques 
with Scene Analysis and Machine Vision, SPIE Proc. No. 1944, 
object completion. 
pages 16-30, Orlando, FL, 1993. 
Petsa, E., and Karras, G.E., Constrained Line-Photogrammetric 
3D-Reconstruction from Stereopairs. IAPRS, vol. XXXIII, 
Amsterdam, 2000. 
Shashua, A., and Kaminski, J.Y., On calibration and 
reconstruction from planar curves. Proc. of the European 
Conference on Computer Vision (ECCV), June 2000, Dublin, 
Ireland. 
Tsai, R.Y., and Thomas, S.H., Estimation Three-Dimensional 
Motion Parameters of a Rigid Planar Patch. 1981. IEEE Trans. 
Acoust., Speech. Sign. Process., 29:1147-1152. 
Tsai, R.Y., Thomas, S.H., and Zhu, W.L., Estimation Tree- 
Dimensional Motion Parameters of a Rigid Planar Patch II: 
Singular value decomposition. 1982. IEEE Trans. Acoust., 
Speech, Sign. Process, 30:525-534. 
Zalmanson, H.G., Hierarchical Recovery of Exterior Orientation 
from Parametric and Natural 3-D Curves. 2000. IAPRS Vol. 
XXXII, Amsterdam. 
   
      
     
  
   
  
   
    
    
  
  
   
  
   
  
  
  
  
   
  
   
  
  
  
  
  
  
  
  
  
   
   
  
  
   
  
  
   
  
  
  
  
   
   
  
  
   
   
   
   
  
  
  
    
  
  
  
  
   
BUND 
* UFPR, Fed 
" UCalgary, D 
KEY WORDS: Ph 
ABSTRACT: 
The use of non-m 
terrestrial surfaces. 
indicating the stab; 
scanning is an exce 
Although the Lasei 
digital image, beca: 
atractive option for 
from photogramme 
processed using a 
methodology to pe 
were taken with a | 
about 2 km? of the 
750 meters and the 
was about 1000 me 
Development LAC" 
of fifteen centimet: 
evaluation of the pr 
New technologies, 
and electronic area: 
photogrammetry. I 
on sophisticated equ 
are now being 
photogrammetric | 
Likewise, Global Pc 
operations necessar 
conducting photogr 
Laser Scanning syst 
obtain altimetric in 
accuracy levels reac 
altimetry and the d 
and Linderberger 1€ 
constituted of GPS, 
LASER scanning « 
Scanner system enat 
of the points which 1 
à group of points iri 
base of a Digital I 
availability of data e 
accuracy in definiti 
objects elevation on 
procedures. On the 
spatial resolutions ar 
of digital images f