osium on 
May 1974 
ry", Final 
gineering 
1al Report 
Eckart Michaelsen 
  
PROBLEMS IN GEOMETRIC MODELLING AND PERCEPTUAL GROUPING OF MAN-MADE 
OBJECTS IN AERIAL IMAGES 
Eckart MICHAELSEN, Uwe STILLA 
FGAN-FOM Research Institute for Optronics and Pattern Recognition 
D 76275 Ettlingen, Germany 
{mich.usti} @fom.fgan.de 
  
Working Group III/4 
KEY WORDS: Image understanding, Semantics, Knowledge representation, Object recognition, Urban objects. 
ABSTRACT 
Structural approaches of pattern recognition are frequently proposed for man-made objects in aerial images. Especially 
model based methods are considered using geometric, topologic and structural knowledge by means of polyhedral 
parametric or generic models. We review some aspects of perceptual grouping with respect to these recognition tasks. 
Emphasis is put on some severe problems encountered in the application. No new results or methods are reported. 
Instead a qualitative comparison and discussion is given between two such approaches with respect to these problems. 
Also the importance of associative access techniques is stressed. 
1 INTRODUCTION 
Research on automatic object recognition in satellite and aerial images has been published for several decades now. 
Especially multi-spectral pixel classification has reached significant maturity, whereas non-local recognition after 
decades of research still remains a challenge. 
1.1 Model based Approaches 
For man-made objects like roads, buildings or vehicles frequently model based approaches have been proposed. Models 
are described in terms of relations of parts forming hierarchical arrangements. Additionally to the part-of relation other 
relations such as adjacency, parallelity e. c. are used. The recognition process then consists of corresponding sub-steps 
of perceptual groupings, searching for constructions consistent with the model. 
1.2 Applications 
This approach has been proposed for quite different recognition tasks. Detection (i. e. searching for special objects), 
classification (i. e. assigning object class labels to data objects) and even reconstruction of all objects in a scene have 
been investigated. Data may have been acquired by different sensors (e. g. visible light, IR, SAR or laser range data). 
Perspectives vary from very oblique to perpendicular. Objects of interest may be fixed in position (e. g. buildings) as 
well as moving around (e. g. vehicles or containers). Buildings may have numerous variations in shape whereas 
vehicles may be categorised into quite homogenous classes. Civil tasks differ frequently from military tasks in the type 
of objects, sensors and perspectives used. This implies distinct object descriptions and strategies of modelling and 
searching. 
1.3 Difficulties 
The euphoric activism of the 80s concerning the application of model based approaches to automatic photogrammetry, 
remote sensing and cartography has decreased during the last years. This results not from a saturation process settling 
the research on secure ground and common views. On the contrary, there seems to be resignation because of practical 
problems with structural models and lack of success in urban scenery. This paper discusses typical problems in the 
construction and application of such grouping approaches to man-made object recognition in aerial images. 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B3. Amsterdam 2000. 877