For the prevention of vandalism video surveillance is increasingly used to monitor places and buildings. Thereby similar problems arise as with standard video surveillance in different areas: An increasingly number of video streams displayed on a huge number of monitors, which exceeds the attention capacity of the security personnel and which makes the manual viewing of the recorded data very time consuming and error-prone.
Within the scope of the project methods will be developed for the automatic analysis of the observation data to support the security personnel and to enable best possible usage of the data. Additionally to the visual information source also vibro-acoustic sensors are used to obtain complementary information and thereby to improve the analysis results and to detect relevant events. An application example for the developed detectors is the automatic detection of vandalism against ticket or cash machines. Visible changes caused by graffiti as well as unusual vibrations and acoustics caused by an unauthorized access or damage will be detected immediately.