Video surveillance systems are generally not regarded as factors which can improve the public subjective impression of safety. One of the main reasons for this lies buried in the fact that commercially available system are still not smart enough to act as reliable support systems for government agencies or security assets for detection and prevention of security/safety related events. They are in fact only used as recording systems used as post-event analysis and fact finding tools in case an event has occurred.
The aim of the project iObserve is to create the basis for future surveillance systems of the next generation. Such systems will be able to combine single low level detections into a more meaningful understanding of a situation on a semantic level - towards the real understanding of a potentially dangerous situation. An important aspect of such a system is its maintenance, configuration and daily operation. The rapidly increasing complexity of the software modules behind it must be translated into actions and commands which can be understood by a knowledgeable but otherwise untrained operator.
Algorithms for the detection of objects, specifically pedestrians, vehicles and baggage items are the grounds on which the iObserve is based on. Software modules already available in the labs of the project partners have been evaluated towards their applicability for iObserve.
DI Martin Forster, Center Communication Systems GmbH, Abteilung Development Image Processing
JOANNEUM RESEARCH Forschungsgesellschaft mbH, Institut für Informationssysteme & Informationsmanagement
Center Communication Systems GmbH
Abteilung Development Image Processing
DI Martin Forster
Tel: +43 (0)1 90 199 - 1304