The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part Bib. Beijing 2008 
667 
• Calculation of x”, y” image coordinates for the PTZ- 
camera with the collinearity equation and initial 
exterior orientation parameters. 
• Calculation of rotation angles a and p with 
a = atan(x”/ c) and P = atan(y” /c). 
plane. 
3.4 Analysis of high-resolution images 
The high-resolution images are stored with time and position 
stamps. A unique assignment of the attributes to the trajectories 
is feasible, because of the observation of stand-alone 
pedestrians. Since there is no automatic classification 
implemented up to now, the pedestrians are interactively 
classified by a human operator. 
4. OUTLOOK 
The two-camera system presented in this work is an essential 
component for the assessment of the quality of shop-locations 
in inner cities, since it delivers both the trajectories of the 
pedestrians and information about additional attributes. So far, 
the project is just in its starting phase. Actual investigations 
concentrate on the improvement of speed and accuracy of the 
control for the exterior orientation and on the zoom-control of 
the PTZ-camera. This will improve the reliability of the 
assignment of the classified pedestrians to their trajectories. 
Furthermore the choice of a person from a number of detected 
individuals up to now is arbitrary. A knowledge-based camera 
control system will ensure the single acquisition of each 
individual or the multiple acquisition of the same person if 
wanted. 
Future work will investigate the opportunity for automatic 
classification of the pedestrians in the high-resolution images. 
For this purpose different classifiers (e.g., AdaBoost, SVM) 
will be evaluated. A training data set is gained by acquisition of 
the high-resolution images. 
The integration of a complete 3D-model of the scene allows an 
improved modelling of the behaviour of pedestrians at obstacles 
or in case of occlusions. Moreover motion models that specify 
the probability for a change in direction or speed at certain 
position will be integrated. 
ACKNOWLEDGEMENT 
Part of the work presented here was carried out in a student 
research project by (in alphabetical order): Jonas Bostelmann, 
Steven Piorun, Falko Schindler, Axel Schnitger, Aiko Sukdolak, 
Martin Wiedeking, 
REFERENCES 
Antonini, G., Venegas Martinez, S., Bierlaire, M., Thiran, J. P., 
2006. Behavioral Priors for Detection and Tracking of 
Pedestrians in Video Sequences. International Journal of 
Computer Vision, 69, p. 159-180. 
Baluja, S., Rowley, H., 2007. Boosting Sex Identification 
Performance. International Journal of Computer Vision, 71, 
p. 111-119. 
Comaniciu D., Ramesh, V., Meer, P., 2003. Kernel-Based 
Object Tracking. IEEE Transactions on Pattern Analysis and 
Machine Intelligence, 25, p. 564-577. 
Lanitis, A., Draganova, C., Christodoulou, C., 2004. Comparing 
Different Classifiers for Automatic Age Estimation. IEEE 
Transactions on Systems, Man and Cybernetics - Part B: 
Cybernetics 34, p. 621-628. 
Leibe, B., Seemann, E., Schiele, B., 2005. Pedestrian Detection 
in Crowded Scenes. IEEE Computer Society Conference on 
Computer Vision and Pattern Recognition (CVPR'05), 
p. 878-885. 
Li, L., Huang, W., Gu, I.Y.H., Tian, Q., 2003. Foreground 
Object Detection from Videos Containing Complex 
Background. Proceedings of the eleventh ACM international 
conference on Multimedia,p. 2-10. 
Moeslund, T. B., Hilton, A., Kruger, V., 2006. A Survey of 
Advances in Vision-based Human Motion Capture and Analysis. 
Computer Vision and Image Understanding, 104, p. 90-126. 
Rosales, R., Sclaroff, S. 1998. Improved Tracking of Multiple 
Humans with Trajectory Prediction and Occlusion Modeling. 
IEEE Conference on Computer Vision and Pattern Recognition. 
Workshop on the Interpretation of Visual Motion. 
Stauffer, C., Grimson, W. E. L., 1999. Adaptive Background 
Mixture Models for Real-time Tracking. IEEE Computer 
Society Conference on Computer Vision and Pattern 
Recognition, 2, p. 246-252. 
Viola, P., Jones, M., 2001. Robust Real-time Object Detection. 
Second International Workshop on Statistical and 
Computational Theories of Vision - Modelling, Learning, 
Computing and Sampling. 
Viola, P., Jones, M. J., Snow, D., 2005. Detecting Pedestrians 
Using Patterns of Motion and Appearance. International 
Journal of Computer Vision, 63, p. 153-161. 
Yilmaz, A., Javed, O., Shah, M., 2006. Object Tracking: A 
Survey. ACM Computing Surveys, 38, Article 13. 
Zhao, T., Nevada, R., 2004. Tracking Multiple Humans in 
Complex Situations. IEEE Transactions on Pattern Analysis 
and Machine Intelligence, 26, p. 1208-1221.