renzhi

Technicians installing camera and flash light.

In the last two years, a network of tracking cameras, with a central computing centre, is being built out in Shanghai. Major arteries are covered, including highways, intersections, and even parking lot entrances of some shopping centres.

Each lane of a road has been assigned a dedicated camera and a flash light, along with a motion detector. Two-way roads have two sets of cameras installed, each set being responsible for the traffic on each direction. As soon as the motion detector senses a car approaching, the flash light is activated and a snapshot is taken. At some key intersections or highways, multiple snapshots of the car are taken. The flash light is quite annoying to drivers, as it is too glaring, especially during night time.

The snapshots are then transmitted to a central repository through a high-speed communication network. In the computing centre, OCR software is used to decode the car license plate, and the information is used for further processing and analysis.

Technicians installing and troubleshooting communication modules.

With that information, it is quite simple to mash up with a map system, such as Google Map, and track every single car in town in real time. Combined that with the geographic data, user profile from the vehicle registration department, profile from the insurance company, driving pattern of each driver, weather and events information in town, and a little bit of analysis and profile on a cluster of cheap computers, this will be a powerful tracking system.

Just add face recognition into the mix, and this would be quite a scary network.