![](bg8.png)
Figure 8: Crowd monitoring example. The expected circu-
lation way has been represented with the green arrows.
tection/matching, to add more efficient temporal blur-
ring in order to have event warning instead of image
warning, to make a discussion over the best type of
kernel to use and finally to compute an adaptative de-
tection threshold C in relation to the learnt distribu-
tion. Moreover, the system only detects anomalies
linked to motion. Expansion to motionless or micro
movement detection is also expected for future in or-
der to enlarge the types of anomalies to detect (person
on the ground, suspect parking ...).
This method has been designed in order to be as
generic as possible with no supervising, as a conse-
quence it may also be used in completely different
fields than video surveillance. We can imagine to use
this approach for detecting perturbations on satellite
imaging of solar streams for example.
REFERENCES
Andrade, E. L., Blunsden, S., and Fisher, R. B. (2006).
Hidden markov models for optical flow analysis in
crowds. In International Conference on Pattern
Recognition (ICPR), volume 1, pages 460–463.
Barron, J., Fleet, D., Beauchemin, S., and Burkitt, T.
(1992). Performance of optical flow techniques.
Computer Vision and Pattern Recognition (CVPR),
92:236–242.
Black, M. J. and Anandan, P. (1996). The robust estima-
tion of multiple motions: parametric and piecewise-
smooth flow fields. Computer Vision and Image Un-
derstanding (CVIU), 63(1):75–104.
Blunsden, S., Andrade, E., and Fisher, R. (2007). Non para-
metric classification of human interaction. In Pattern
Recognition and Image Analysis (PRIA).
Brostow, G. j. and Cipolla, R. (2006). Unsupervised
bayesian detection of independant motion in crowds.
In Computer Vision and Pattern Recognition (CVPR).
Burges, C. J. C. (1998). A tutorial on support vector
machines for pattern recognition. Data Mining and
Knowledge Discovery, 2:121–167.
Dong, L., Parameswaran, V., Ramesh, V., and Zoghlami, I.
(2007). Fast crowd segmentation using shape index-
ing. In International Conference on Computer Vision
(ICCV).
Duda, R., Hart, P., and Stork, D. (2001). Pattern Classifica-
tion. John Wiley & Sons Inc.
Figueiredo, M. and Jain, A. (2002). Unsupervised learning
of finite mixture models. Pattern Analysis and Ma-
chine Intelligence, 24(3):381–396.
Freund, Y. and Schapire, R. E. (1996). Experiments with
a new boosting algorithm. In Proceedings of the
30th International Conference on Machine Learning
(ICML), pages 148–156.
Han, B., Comaniciu, D., Zhu, Y., and Davis, L. S. (2008).
Sequential kernel density approximation and its appli-
cation to real-time visual tracking. Pattern Analysis
and Machine Intelligence, 30:1186–1197.
Harris, C. and Stephens, M. (1988). A combined corner
and edge detector. In Alvey Vision Conference, pages
147–151.
Hu, W., Xiao, X., Fu, Z., Xie, D., Tan, T., and Maybank,
S. (2006). A system for learning statistical motion
patterns. IEEE Pattern Analysis And Machine Intelli-
gence, 28:1450–1464.
Junejo, I. N. and Foroosh, H. (2007). Trajectory rectifica-
tion and path modeling for video surveillance. In In-
ternational Conference on Computer Vision (ICCV).
Li, Y. and Ai, H. (2007). Fast detection of independent mo-
tion in crowds guided by supervised learning. In In-
ternational Conference on Image Processing (ICIP).
Proesmans, M., Gool, L. V., Pauwels, E., and Oosterlinck,
A. (1994). Determination of optical flow and its dis-
continuities using non-linear diffusion. In European
Conference on Computer Vision (ECCV).
Rabaud, V. and Belongie, S. (2006). Counting crowded
moving objects. In Computer Vision and Pattern
Recognition (CVPR).
Sidla, O. and Lypetskyy, Y. (2006). Pedestrian detection
and tracking for counting applications in crowded sit-
uations. In Advanced Video and Signal Based Surveil-
lance (AVSS).
Singh, S., Machiraju, R., and Parent, R. (2007). Visual anal-
ysis of trajectory clusters for video surveillance.
Stauffer, C. and Grimson, W. E. L. (2000). Learning pat-
terns of activity using real-time tracking. IEEE Pat-
tern Analysis And Machine Intelligence, 22:747–757.
Tipping, M. E. (2001). Sparse bayesian learning and the rel-
evance vector machine. Journal of Machine Learning
Reasearch 1.
Velastin, S. A., Boghossian, B. A., and Vicencio-Silva,
M. A. (2006). A motion-based image processing
system for detecting potentielly dangerous situations
in underground railway stations. Transportation Re-
search Part C: Emerging Technologies, 14.
Zhao, T. and Nevatia, R. (2004). Tracking multiple humans
in complex situations. IEEE Pattern Analysis And Ma-
chine Intelligence, 26:1208–1221.
AN UNSUPERVISED LEARNING BASED APPROACH FOR UNEXPECTED EVENT DETECTION
513