inbul 2004 
(9) 
(10) 
pH 
om]. 
(12) 
Y,) 
x.) 
nputed by 
is started. 
of stereo 
de 
ition. To 
follows: 
(13) 
Ay) 
and Ay 
computed 
(14) 
v, Ay), and 
y of this 
    
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004 
  
So, we obtain the modified algorithm with the computational 
complexity decreased in 2a times relative to traditional scheme. 
It is also important that the algorithm described can be easily 
implemented at any parallel architecture. 
4. CONCLUSIONS 
The problem of increasing of computational speed of automatic 
stereo matching is an important issue for designing low cost 
photogrammetric systems on PC-compatible computers. In this 
paper the solution for this problem is proposed based on image 
informative characteristics and non-traditional implementation 
of computations in a sliding correlation window. 
Both useful described techniques allows obtaining the economy 
of computational time up to 20-25 times while processing of 
typical Russian TK-350 spaceborn images, and could be used 
for another data processing schemes. 
REFERENCES 
Sibiryakov A., 2000. DTM generation from Russian TK-350 
space imagery in the PC-based photogrammetric system Z- 
Space, International Archives of Photogrammetry and Remote 
Sensing, Vol. XXXII, Part B4, Amsterdam. 
Huang T.S., 1983. Image Sequence Processing and Dynamic 
Scence Analysis. — Springer-Verlag, New York.