Towards self-organising Smart Camera Systems

Abstract

Smart Camera Systems consist of large numbers of net-worked cameras which can adjust their fields of view by panning, tilting and zooming. Each Smart Camera is an embedded systems that does not just capture raw video streams but also analyses video data locally using computer vision techniques. Apart from image processing tasks the second important issue is the management of such systems, e.g. camera alignment and calibration. Due to the increasing number of cameras in these systems manual administration becomes hardly feasible. Therefore, algorithms for autonomous system organisation are needed. As a basis for these algorithms, in this paper we propose a distributed system architecture which is tailored to the requirements of Smart Camera Systems. Inspired by self-organisation -a major paradigm of Organic Computing- this paper presents an algorithm relying on our distributed architecture for one important management problem in Smart Camera Systems, i.e. the spatial partitioning of an area under observation. © 2008 Springer-Verlag Berlin Heidelberg.