Tracking People from a Camera Network 

We are interested in developing a people tracking system that can track the target robustly through a highly unstructured indoor environment (with clutter, multiple humans, and unconstrained lighting conditions) using a network of cameras and multiple cues. The following figure illustrates the concept of the proposed system. The system consists of a network of omnidirectional and perspective video cameras. This combination will minimize possibilities of losing track of the target when it is occluded by other people and/or objects. A perspective video camera has a limited viewing angle and is directional. The human target could be occluded from multiple normal video cameras if he/she is among crowd. An omnidirectional camera mounted on the ceiling, on the other hand, has 360 degree viewing angle and, by virtue of being above the "action," is less prone to problems of occlusion.  

Related Publications:

• D. Chen and J. Yang, Online Learning Region Confidences for Object Tracking, The Second Joint IEEE International Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveillance, in conjunction with the Tenth IEEE International Conference on Computer Vision (ICCV'05), 2005.
• W. Wang, D. Chen, W. Gao and J. Yang, Modeling Background from Compressed Video, The Second Joint IEEE International Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveillance, in conjunction with the Tenth IEEE International Conference on Computer Vision (ICCV'05), 2005.
• R. Stiefelhagen, X. Chen, J. Yang, Capturing Interactions in Meetings with Omnidirectional Cameras, International Journal of Distance Education Technologies , Vol. 3, No.3, pp. 34-47, 2005.
• D. Chen and J. Yang, Image Registration in an Uncalibrated Omnidirectional Camera Network, In Proceedings of the WACV 2005, pp. 427-432, 2005.
• D. Chen, R. Malkin, J. Yang, Multimodal Detection of Human Interaction Events in a Nursing Home Environment, Proceedings of ACM International Conference on Multimodal Interface (ICMI), pp. 82 - 89, 2004.
• X. Chen, J. Yang, A. Waibel, Calibration of a Hybrid Camera Network, Proceedings of ICCV 2003 , pp. 150-155, 2003.
• X. Chen and J. Yang, Towards monitoring human activities using an omnidirectional camera, Proceedings of ICMI 2002, pp. 423 - 428, 2002.