Full text: Close-range imaging, long-range vision

  
+. — 
19% century. 
  
facades, like 
fferent points 
fferent image 
to different 
Jiscontinuous 
entioned. But 
due to one or 
fferent points 
to determine 
ight. 
to be used in 
a hierarchical 
start from a 
imum positive 
; data may be 
el. 
images which 
LSM) (Helava 
s usually not a 
asurement of a 
per image. At 
yuld allow the 
S. 
nd approximate 
arch areas, the 
  
  
  
  
  
  
  
  
   
    
  
  
  
Figure 3: 
    
p 
nts re-projected into a oriented 
i 
Geodetic measured poi 
image. 
  
  
   
  
determination of sufficiently exact initial values for the 
parameters of the least squares matching and the evaluation of 
results due to ray intersection results. A search process looks for 
the best image showing the work space in the largest resolution 
and most possible perpendicular. This image is automatically 
used to provide the master matrix in the matching process. All 
other images are evaluated due their suitability as matching 
partners. This can be calculated via the co sinus of the angle 
between the normal vector of the facade and the image 
acquisition vector (known from the orientation data) 
  
Bass 
OCSCRE EGRE: XM 
adient images 
  
Fig: 5: Four gr 
2.2 Region Growing 
The tie and control points are used as seeds for region growing 
algorithms. It can be assumed, that the region close to the tie 
points lies on the same object plane as the tie point itself. As 
information on the surface also some points measured with a 
reflector-less total station may be used (Fig. 3). 
LIES 
    
Ce 
  
Riche pid sav Es ses 
E SX den: 
Sowg EI t" quoe Huc ura] Meine Fuscus Er ms re rms. (GU Ss 
CECI ET 
Tub exter ADM. Mola Potoqocnete Hh 
     
  
NN E 
  
  
ice GS 
PT 
É SHOE URN HE A as 
]uitees m ee as iris ITE morse Anse 
Figure 4: Matching candidates projected into two images 
using orientation data of a first iteration step of the 
redion growing algorithm. 
Therefore the same initial values for the points in the 
neighbourhood may be used after they have been projected into 
the oriented images (Fig. 4). In the future we will use an interest 
operator to search for a good matching candidate in a suitable 
distance and direction from the tie point using. 
  
—481— 
  
 
	        
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.