Multi-camera people tracking using Bayesian networks

A multi-camera tracking system is considered for video surveillance, where the cameras have non-overlapping views and the system is required to be robust under different lighting conditions. As people enter the scene, they are segmented out as foreground objects. Facial features, texture and color features of clothing, and likely location of people are extracted and matched with like data stored in a database that is dynamically constructed. The top 3 matches are used in a Bayesian network, together with a face detection confidence measure, and the likelihood of a person's location for inferencing. To establish identity, results over 16 consecutive frames are used in a majority voting scheme. When the person is identified, the extracted data is stored in the database for future use. To assess the performance of the system, experiments were conducted on a database of 11 people, simulating 4 different camera views. Identification accuracies in excess of 95.45% were obtained in different experiments.

[1]  Ramin Zabih,et al.  Bayesian multi-camera surveillance , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[2]  Michael Isard,et al.  Contour Tracking by Stochastic Propagation of Conditional Density , 1996, ECCV.

[3]  Robert C. Bolles,et al.  Background modeling for segmentation of video-rate stereo sequences , 1998, Proceedings. 1998 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No.98CB36231).

[4]  Seiji Inokuchi,et al.  CAD-based object tracking with distributed monocular camera for security monitoring , 1994, Proceedings of 1994 IEEE 2nd CAD-Based Vision Workshop.

[5]  John A. Marchant,et al.  Colour Invariance at a Pixel , 2000, BMVC.

[6]  Jake K. Aggarwal,et al.  Tracking Human Motion in Structured Environments Using a Distributed-Camera System , 1999, IEEE Trans. Pattern Anal. Mach. Intell..

[7]  Stan Sclaroff,et al.  Improved Tracking of Multiple Humans with Trajectory Predcition and Occlusion Modeling , 1998 .

[8]  Shaogang Gong,et al.  Tracking Multiple People Under Occlusion Using Multiple Cameras , 2000, BMVC.

[9]  Y. V. Venkatesh,et al.  An integrated automatic face detection and recognition system , 2002, Pattern Recognit..

[10]  Finn Verner Jensen,et al.  Introduction to Bayesian Networks , 2008, Innovations in Bayesian Networks.

[11]  Stuart J. Russell,et al.  Object Identification: A Bayesian Analysis with Application to Traffic Surveillance , 1998, Artif. Intell..

[12]  Alexander H. Waibel,et al.  Skin-Color Modeling and Adaptation , 1998, ACCV.

[13]  Paolo Remagnino,et al.  Multi-camera colour tracking , 1999, Proceedings Second IEEE Workshop on Visual Surveillance (VS'99) (Cat. No.98-89223).