Multi-camera image measurement and correspondence

This paper describes part of a project to develop a multi-camera image tracking system. The system will be developed in the context of a surveillance and monitoring task. principally targeted at tracking people through indoor and outdoor environments. Moving objects were detected for each camera viewpoint using background subtraction. Correspondence between viewpoints is then established by epipole line analysis. Measurements were made in 3D on each corresponded object using a least squares estimate. The error covariance of each measurement was computed using a sensory uncertainty field. This information will be used to interpret the behaviour of the observed object and learn the relationships between each camera viewpoint and the scene under observation. (C) 2002 Elsevier Science Ltd. All rights reserved.