JAIN, R.; KASTURI, R; SCHUNCK B.G.,1995. Machine 
Vision. McGraw-Hill, Inc. Singapore. 549p. 
LUGNANI, J.B., 1988. Introduction to Phototriangulation. 
Curitiba: ed UFPR, 134p. 
MERCHANT, D.C., 1988. Analytical Photogrammetry: Theory 
and Practice. Part 1, Columbus, Ohio, Department of Geodetic 
Science, The Ohio state University. 
MIKHAIL, E.M.; BETHEL, J; McGLONE, J.C. 2001. 
Introduction to Modern Photogrammetry. 479p. 
SONY, 2002. Digital Still Camera Cybershot DSC-F717. Sony 
Corporation, 121p. 
WEVER C., LINDEBERGER J., 1999. Experinces of 10 years 
laser scanning. www ifp.uni-stuttgart.de/publications/ phowo99 
/wever.pdf, 125-132p. 
http//www.optech.on.ca (accessed 5/25/2003) 
Appendix — Triangulation data 
Program: BundleH 
Site: Centro Politecnico - UFPR - Curitiba - Parana - Brazil. 
Camera : Sony DSC-F717 - resolution 2560 x 1920 pixel 
Calibration information: ( unit = pixel ) 
Focal Length :2931.722 
Coordenate x of Principal Point : -71.648 
Coordenate y of Principal Point : -40.965 
Symmetric Radial Distortion Coefficients: 
K1 = -2.63950e-008 
K2 = 3.24240e-015 
K3 = 3.06100e-022 
Decentric Distortion Coefficients: 
P1 = -4.13010e-007 
P2 = 2.42890e-007 
Affine Deformation Parameters: 
A = -1.34550e-004 
B = -2.03270e-005 
Photogrammetric Refraction 
Atmospheric Refraction e45 Coefficient : 9.32533e-006 
(Saastamoinen Model) 
Flight information : 
Average terrain height above sea level : 909.394 m 
Average flying height above sea level : 1648.583 m 
Average flying height above terrain :°739 189 m 
Number of Images 248 
Pixel Ground Sample Distance 5 :0,252.m 
Adjustment input: 
Ground Points Observations Standard Deviation : 0.250 m 
Image Observations Standard Deviation : 1.0 pixel 
Sigma Apriori (a priori variance of unit weight): 1.0 
Number of Parameters (EO) = 6 x n' > of images : 78 
Coordinates = 3 x n° of points : 105 
Observation Equations = 2 x n° of observ ations : 190 
Number of Constraints Equations :-90 
: 97 
Redundancy 
Processing 
Iteration = 15 
Convergency = 0.000001 
SigmaPosteriori = 0.196674 
Global Test — significance 5% 
OK if Q2table >= Q2computed 
Q2computed 19.08 
Q2table 30.24 
Passed. 
Image Observations Residues 
V(Rxy) 0.362 (pixel) Mean 
V(Rxy) 0.204 (pixel) Standard Deviation 
V(Rxy) 0.919 (pixel) Max Resultant 
V(Rxy) 0.040 (pixel) Min Resultant 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004 
   
  
   
   
    
    
  
   
  
   
  
  
  
  
   
   
    
  
    
   
   
  
  
  
  
  
  
  
  
   
   
  
    
   
   
   
   
   
   
    
  
  
   
  
   
  
    
     
  
  
    
   
   
   
  
  
     
Bundle Adju 
KEY WORDS: B 
ABSTRACT: 
In this paper we in 
refining a set of par 
a cost function. In 1 
function measures | 
architectural or terr 
within particular sc 
linear features. Usir 
and the detected im: 
Second, we study t| 
images. The seconc 
the theoretical boun 
segments can be rel 
Accurate parameter 
based applications. 
tion are intimately li 
the knowledge of th 
structure reconstruc 
matching can provic 
mation. For instance 
that both the compu 
try community have 
communities are cui 
estimation of nume 
On the one hand, li 
tive geometry provi 
(Heuel, 2001). Hen 
rameters (Faugeras € 
Davidson, 2003) are 
nity. These robust es 
pretation. On the otl 
metric community c 
the viewing parame 
estimation of the pai 
a refinement of the : 
((Triggs et al., 200( 
1993)). Usually, on 
man-made environm 
5S0, it is of importan 
process, as well as t 
cation can be done | 
each parameter. If th 
are Gaussian, we can 
sociate an uncertaint 
between different pai 
2000), (Hartley and ? 
2. 
2.1 Introduction 
Bundle adjustment i 
set of parameters suc 
point and line positi 
function based on all 
trol Points (GCP), tie