The estimation of sun emission characteristics is not 
required. 
9 The desired contours are tested to improve the 
matching results. 
o The shadows' evidences can be taken in account 
immediately in the matching process. 
The last point is clear from the look at the "shape" of any 
typical image of a house with a shadow. We can account 
the shadow casting at the stage 1 of our matching 
algorithm (while creating the "shape"). Additionally, we 
can especially check that the shadow region of the mid- 
level approximation is dark enough. 
The following improvement of this technique is connected 
with the account of non-pixel events. For example, one can 
extract the straight lines' segments on contours and 
consider them as contour events. 
4. CONCLUSION 
A new approach for model-based image analysis named 
the Events-based image Analysis (EA) is proposed. From 
EA point of view, any certain procedure of image 
understanding can be interpreted as a procedure of 
evidence fusion. Any fact about the whole image, about its 
part or even about one proper pixel can be the evidence, 
and the any proposition about the scene observed is the 
hypothesis that requires to be proved or escaped based on 
these evidences. In this paper the EA formalism was 
outlined in the Bayesian terms. 
This approach allows to compose the power of sample- 
based methods and the flexibility of model-based methods 
without the direct comparison of objects or images. The 
most important properties of EA procedures that are 
principally improper for the comparison-based techniques 
are the following: 
e  theusage of generic models; 
e the usage of hierarchical models; 
e the usage of non-homogeneous information. 
Based on EA ideas the complex technique for house 
detection was proposed. It provides the easy fusion of 
contour and intensity information for 3D-model validation. 
This technique was realized as a low cost application on 
IBM PC and preliminary tested using a set of images of 
Ufa city (Russia). The preliminary results demonstrate that 
the house detection is satisfactory enough. 
The future work will be connected with the following 
improving and testing of this algorithm. 
902 
   
References: 
i 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B3. Vienna 1996 
Stepanov A, Zheltov S. and Visilter Yu., "Shape 
analysis using Pytiyev morphologic paradigm and 
its' use in machine vision". - SPIE vol. 2350, 1994. 
Houle D. and Malowany A., "Object recognitions 
using fast Adaptive Hough transform", IEEE Pacific 
Rim Conference on Communications, Computers 
and Signal Processing, June, 1-st-2nd, 1989. 
Huertas A, Bejanin M. and Nevatia R., "Model 
Registration and "Validation". - in "Automatic 
extraction of man-made objects from aerial and 
space images"/ edby A.Gruen at al, Basel: 
Birkhauser, 1995. 
Muller W. and Olson J., "Automatic matching of 
3D-Models to Imagery". - in "Automatic extraction 
of man-made objects from aerial and space images"/ 
ed.by A.Gruen at al, Basel: Birkhauser, 1995.