OPEN ACCESS
The research addressed in this paper consists in developing an automatic system for detecting/reidentifying moving people in different sites with non-overlapping views. Firstly, we propose an automatic process for silhouette extraction. Then, a new colour descriptor is proposed for characterizing the silhouettes of the detected people. Once all the appearance information of each passage is extracted, a graph-based algorithm is used to realize the comparison of passages of people in front of cameras and to make the final decision of reidentification. The global system is tested on two real and difficult data sets recorded in very different environments. The experimental results show that our proposed system leads to very satisfactory results.
RÉSUMÉ
Les travaux de recherche présentés dans cet article ont pour objet le développement d'un système automatique multicaméra de détection/réidentification de personnes évoluant dans différents sites surveillés. Nous proposons dans un premier temps un processus automatique d'extraction de l'individu du fond permettant d'avoir des silhouettes de bonne qualité. Par la suite, un nouveau descripteur est proposé afin de caractériser les silhouettes de plusieurs personnes. Une approche basée sur l'analyse spectrale de graphes de similarités est proposée pour réaliser la réidentification. Deux bases de données représentant le passage de plusieurs personnes devant des caméras ont été utilisées pour l'évaluation des algorithmes. Les résultats expérimentaux très satisfaisants permettent de valider la robustesse de l'approche proposée.
surveillance system, person detection, colour-based descriptor, person reidentification, colour invariant, dimension reduction
MOTS-CLÉS
vidéosurveillance, détection de personnes, réidentification de personne, signature colorimétrique, invariant couleur, réduction de dimension
Ballard D., Brown C. (1982). Computer Vision, Prentice Hall Professional Technical Reference.
Belkin M., Niyogi P. (2003). « Laplacian eigenmaps for dimensionality reduction and data representation », Neural Computation, vol. 15, n° 6, p. 1373-1396.
Birchfield S., Rangarajan S. (2005). « Spatiograms versus Histograms for Region-Based Tracking », IEEE Computer Society Conference on Computer Vision and Pattern Recognition, vol. 2, p. 1158-1163.
Buchsbaum G. (1980). « A spatial processor model for object color perception », Journal of the Franklin Institute, vol. 310, n° 1, p. 1-26.
Elgammal A., Harwood D., Davis L. (2000). « Non-parametric model for background subtraction », Proceedings of the 6th European Conference on Computer Vision, p. 751-767.
Finlayson G., Hordley S., Schaefer G., Yun Tian G. (2005). « Illuminant and device invariant colour using histogram equalisation », Pattern Recognition, vol. 38, n° 2, p. 179-190.
Fisher R. (1936). « The use of multiple measurements in taxonomic problems », Annals of Eugenics, vol. 7, p. 179-188.
Gevers T., Stokman H. (2004). « Robust Histogram Construction from Color Invariants for Object Recognition », IEEE Transactions on Pattern Analysis and Machine Intelligence, p. 113-118.
Gheissari N., Sebastian T., Tu P., Rittscher J., Hartley R. (2006). « Person Reidentification Using Spatiotemporal Appearance », Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Washington, DC, USA, p. 1528-1535.
Hartigan J. (1975). Clustering Algorithms, John Wiley & Sons, Inc. New York, NY, USA.
Hotelling H. (1933). « Analysis of a complex of statistical variables into principal components », Journal of Educational Psychology, vol. 24, p. 417-441.
Javed O., Rasheed Z., Shafique K., Shah M. (2003). « Tracking across multiple cameras with disjoint views », Ninth IEEE International Conference on Computer Vision.
Javed O., Shafique K., Shah M. (2005). « Appearance Modeling for Tracking in Multiple NonOverlapping Cameras », IEEE Computer Society Conference on Computer Vision and Pattern Recognition, vol. 2, p. 26-33.
Kettnaker V., Zabih R. (1999). « Bayesian multi-camera surveillance », IEEE Computer Society Conference on Computer Vision and Pattern Recognition, vol. 2.
Kim K., Chalidabhongse T., Harwood D., Davis L. (2004). « Background modeling and subtraction by codebook construction », International Conference on Image Processing, vol. 5.
Kullback S., Leibler R. (1951). « On information and sufficiency », Annals of Mathematical Statistics, vol. 22, n° 1, p. 79-86.
Madden C., Piccardi M., Zuffi S. (2007). Comparison of Techniques for Mitigating the Effects of Illumination Variations on the Appearance of Human Targets, vol. 4842 of Lecture Notes in Computer Science, Springer.
Nadler B., Lafon S., Coifman R., Kevrekidis I. (2005). « Diffusion Maps, Spectral Clustering and Eigenfunctions of Fokker-Planck Operators », Advances in Neural Information Processing Systems, p. 955-962.
Nakajima C., Pontil M., Heisele M., Poggio T. (2003). « Full body person recognition system », Pattern Recognition, vol. 36, n° 9, p. 1997-2006.
Ng A., Jordan M., Weiss Y. (2001). « On spectral clustering: Analysis and an algorithm », Advances in Neural Information Processing Systems, MIT Press, p. 849-856.
Parzen E. (1962). « On the estimation of a probability density function and mode », Annals of Mathematical Statistics, vol. 33, p. 1065-1076.
Porikli F., Tuzel O. (2003). « Human Body Tracking by Adaptive Background Models and Mean-Shift Analysis », IEEE International Workshop on Performance Evaluation of Tracking and Surveillance.
Stauffer C., Grimson W. (1999). « Adaptive background mixture models for real-time tracking », Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, vol. 2, p. 246-252.
Truong Cong D.-N., Khoudour L., Achard C., Phothisane P. (2009). « People Re-identification by Means of a Camera Network Using a Graph-based Approach », Proceeding of IAPR Conference on Machine Vision Applications.
von Luxburg U. (2007). « A Tutorial on Spectral Clustering », Statistics and Computing, vol. 17, n° 4, p. 395-416.
Wang X., Doretto G., Sebastian T., Rittscher J., Tu P. (2007). « Shape and Appearance Context Modeling », EEE 11th International Conference on Computer Vision, p. 1-8.
Wren C., Azarbayejani A., Darrell T., Pentland A. (1997). « Pfinder: Real-time tracking of the human body », IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 19, n° 7, p. 780-785.
Yoon K., Harwood D., Davis L. (2006). « Appearance-based person recognition using color/path-length profile », Journal of Visual Communication and Image Representation, vol. 17, n° 3, p. 605-622.
Yu Y., Harwood D., Yoon K., Davis L. (2007). « Human appearance modeling for matching across video sequences », Machine Vision and Applications, vol. 18, n° 3, p. 139-149.