hul 2004 International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004 
  
distortion regions can't be estimated with any fixed-size block- 
matching algorithm, the window warping technique is used in 
block matching process. The position of a reference window to 
warping is adaptively obtained according to the degree of 
reliability calculated by estimates disparities in the previously. 
In order to reduce noise sensitivity and reach higher efficiency 
simultaneously, we used a modified hierarchical block- 
matching algorithm. 
The experimental results show that considerable improvements 
are obtained especially in projective distortion region, and 
texture-mapping result demonstrates the enhanced performance 
of proposed algorithm. 
texture 
References 
RSEN G. Wolberg, Digital Image Warping, IEEE, pp. 153-160, 1994. 
G. Medioni and R. Nevatia, Segment-Based Stereo Matching, 
Computer Vision, Graphics, and Image Processing, vol. 31, no. 
1, pp. 2-18, July, 1985. 
K. P. Han, T. M. Bae, and Y. H. Ha, A Compact Stereo 
Matching Algorithm Using Modified Population-Based 
Incremental Learning, The Journal of The Institute of 
Electronics Engineers of Korea, vol. 36, no. 10, pp. 103-112, 
Oct. 1999. 
P. N. Belhumeur and D. Mumford, A Bayesian Treatment of 
the Stereo Correspondence Problem Using Half-occlusion 
Regions, IEEE Conf. on Computer Vision and Pattern 
Recognition, pp.506-512, 1992. 
98 
T. Kanade and M. Okutomi, A Stereo Matching Algorithm with 
texture an Adaptive Window, IEEE Trans. on Patt. Anal. Machine 
ft view Intell., vol. 16, no. 9, pp. 920-932, Sep. 1994. 
ge, (d) 
Z. F. Wang and N. Ohnish, Intensity-Based Stereo Vision: from 
3-D to 3-D, SPIE, vol. 2354, pp. 434-443, Nov. 1994. 
atching 
(noise) 
curs in 
sv The 
ontains 
enon is 
ised in 
(Figure 
are the 
> result 
ecially 
itching 
uced. 
ed, we 
These 
otating 
ure 13 
ntional 
ility in 
-based 
iprove 
jective 
773