Cooperative surveillance of multiple targets using mutual information
Date
2008
Authors
Sommerlade, E.
Reid, I.
Editors
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Conference paper
Citation
Proceedings of the ECCV Workshop on Multi-camera and Multi-modal Sensor Fusion Algorithms and Applications (M2SFA2), 2008, pp.1-12
Statement of Responsibility
Eric Sommerlade and Ian Reid
Conference Name
Workshop on Multi-camera and Multi-modal Sensor Fusion Algorithms and Applications - M2SFA2 (18 Oct 2008 - 18 Oct 2008 : Marseille, France)
Abstract
This work presents a method to control multiple, but diverse pan-tilt-zoom cameras which are sharing overlapping views of the same spatial location for the purpose of observation of this scene. We cast this control input selection problem in an information-theoretic framework, where we maximise the expected mutual information gain in the scene model with respect to the observation parameters. The scene model yielding this information comprises several dynamic targets, augmented by one which has not yet been detected. The infor- mation content of the former is supplied directly by the uncertainties computed using a Sequential Kalman Filter tracker for the observed tar- gets, while the undetected is modelled using a Poisson process for every element of a common ground plane. Together these yield an information- theoretic utility for each parameter setting for each camera, triggering collaborative explorative behaviour of the system. Overall this yields a framework in which heterogeneous active camera types can be integrated cleanly and consistently, obviating the need for a wide-angle supervisor camera or other artificial restrictions on the cam- era parameter settings.