A Virtual Curtain for the Detection of Humans and Access Control - Advanced Concepts for Intelligent Vision Systems - page 107

Finally, it must be noted that during our tests, we observed that for low global
error rates, the number of FN is considerably lower than the number of FP. In
other words, the system is naturally more inclined to reject a single person than
to allow to a group of two persons to pass the door. For an access control system,
it is a welcomed property.

5 Conclusions

This paper introduces the concept of virtual curtain that is obtained by the
registration of two linear laser scanners that measure distances in a same plane.
Despite intrinsic shortcomings, originated by effects like occlusion or shadowing,
features derived from an object crossing a virtual curtain permit to interpret the
scene. In particular, it is shown how it is possible to differentiate between several
scenarios for the context of access control. Features are first extracted for every
intersection between an object and the curtain, then they are concatenated to
provide a temporal signature. This signature is handled by a classification process
that identifies the ongoing scenario. Results show that a high recognition rate
is achievable for a pre-defined set of training and testing scenarios. In practice,
we will have to wait for international standardization bodies or organizations to
elaborate some criteria to benchmark the performances for a use under variable
operational conditions. But our results proof that our system is tractable and
usable for the interpretation of a scene.

Acknowledgments. This work was supported by the Belgian Walloon Re-

http://www.wallonie.be

gion ( ). S. Piérard has a grant funded by the FRIA

http://www.frs-fnrs.be/ http://www.

( ). We are also grateful to B.E.A .S.A.(

bea.be ), and in particular to Y. Borlez, O. Gillieaux, and E. Koch, and for their

invaluable help.

References

1. Jain, A., Flynn, P., Ross, A.: Handbook of Biometrics. Springer, Heidelberg (2008)
2. Sequeira, V., Boström, G.: Gonçalves, J.: 3D site modelling and verification: usage
of 3D laser techniques for verification of plant design for nuclear security applica-
tions. In: Koschan, A., Pollefeys, M., Abidi, M. (eds.) 3D Imaging for Safety and
Security, pp. 225-247. Springer, Heidelberg (2007)
3. Boulgouris, N., Hatzinakos, D., Plataniotis, K.: Gait recognition: a challenging
signal processing technology for biometric identification. IEEE Signal Processing
Magazine 22(6), 78-90 (2005)
4. Nixon, M., Carter, J., Shutler, J., Grant, M.: New advances in automatic gait
recognition. Elsevier Information Security Technical Report 7(4), 23-35 (2002)
5. Nixon, M.: Gait biometrics. Biometric Technology Today 16(7-8), 8-9 (2008)
6. Nixon, M., Tan, T., Chellappa, R.: Human identification based on gait. Springer,
Heidelberg (2006)
7. Foster, J., Nixon, M., Prügel-Bennett, A.: Automatic gait recognition using area-
based metrics. Pattern Recognition Letters 24(14), 2489-2497 (2003)