Problem being adressed:
Video surveillance in cities today is expensive to build largely because of the cost of cabling and the bandwidth constraints that often make in necessary to place recording servers on poles in the street to be physically close to the cameras. For maintenance purposes, servers on poles in the street are less than ideal. Therefore cost is still a major obstacle to building out large city surveillance projects.
Tracking objects automatically between such recording servers is not possible due to latency issues. A typical situation is that on camera with a birds eye view overlooks an area and other cameras with the ability to zoom in on a suspect are placed at street level and are controlled using information gained from the overview. This requires low latency communication between the components. Currently such a setup will only work inside one building with all components on the same high speed internal network. In city surveillance the latency in a larger network and the associated cost of special purpose networks does not allow the use of such technology.
Use for society
Video surveillance can help to make cities safer and smarter.
Terrorist and criminal activities may be prevented or at least be forensically examined to arrest perpetrators and collaborators. This has been proven in the aftermath of the Boston marathon attack and also in the attacks in Paris and Bruxelles.
Other uses range from improvement of traffic management to city maintenance improvements and cost savings. Obviously care needs to be taken that the civil rights of the population in general are not infringed beyond what is absolutely necessary to achieve the objectives.
Overall objectives
The first objective is to show that video surveillance servers can be placed inside the Metro network. This is expected to yield significant cost savings thus making video surveillance in cities more viable and increasing camera density.
The second objective is to show that the low latency of Metro networks allows the handover of object tracking information between cameras in real time. This will be demonstrated in the Berlin demo.
