Camera networks are increasingly found in the public space, for example at railway stations, airports or even in inner city areas. It is precisely after the violent overlaps of the recent past that surveillance technologies are the focus of politics. However, in order to make the evaluation efficient and manageable, systems must be developed that make the data flow controllable.
The Department VID focuses on the intelligent evaluation of video data in sensor networks. For this purpose, a camera network was installed specifically at IOSB in order to implement and further develop service-oriented system architectures, evaluation methods (for example, personal tracking and identification) and calibration methods. The self-research project N.E.S.T. Forms the basis for this.
The analysis of video sequences in sensor networks ranges from the "classic" scenario of the detection of an abandoned suitcase, to the camera-encompassing recognition of persons.
Left: Goal of modern sensor network monitoring. A single person should already suffice to request specific information or to get it automatically. Right: Classical monitor monitoring - inefficient and associated with high personnel expenses.