The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008 
1211 
Parts of buildings may be occluded by other buildings, problems. 
Figure 7: Shaded visualization of matched DSM with some details, upper right = DMC-image 
7. Conclusion 
Digital surface models in urban areas can be generated based on 
large scale DMC-images, but some difficulties have to be 
expected, such as caused by: radiometric problems, occlusions, 
shadows and vegetation. 
In the area including complex objects such as small and large 
buildings and trees close to roofs, the detection and definition 
of comer points in the building was poor, leading to the 
possibility of errors. 
But we can deduce the following results: 
High accuracy of generated DSM based on DMC-images 
with a standard deviation of the height between 0.8 and 1.2 
GSD. 
Generation of good DSM depends on the image quality 
and the object visibility. 
The matching parameters should be optimized according to 
the characteristics of each area. 
REFERENCE 
Gruen, A.W. and Baltsavias, E.P., 1987: High-precision image 
matching for digital terrain model generation, IAPRS, Vol 25, 
No 3/1: 284-296. 
Heipke, C., 1996: Overview of Image Matching Techniques - 
http://phot.epfl.ch/workshop/wks96/art_3_l .html(January 2008) 
Jacobsen, K., 2006: Digital surface Models of city Areas by 
very High Resolution Space Imagery. EARSeL Workshop on 
Urban Remote Sensing, Berlin March 2006, on CD 
Jacobsen, K., 2007: Manual of program system BLUH, Institute 
of Photogrammetry and Geoinformation, Leibniz University 
Hannover, Germany