Image recognition system

Image recognition system

Details of several surveillance systems have been given in this section together with an outline of the research that has led to the development of an effective image recognition system. This system is so precise that it is capable of identifying an image received from a CCTV system, even if it is taken from transmissions that contain multi-images. In other words, it claims to be able to identify a face in a crowd.

There is now so much confidence in this system that it is being brought into operation at shopping centres in East London. The aim in this pilot scheme is touse a surveillance system that relies on CCTV cameras to recognise the faces of known criminals.

The UK's Metropolitan Police are running a six-month pilot scheme with the local authority of Newham. The developers believe that the system will help them track criminals and prevent crime. Civil rights workers are not, however, convinced and believe that it intrudes on an individual's privacy.

The system in use is called"Mandrake" and has already been used in the USA but not in Europe. The system scans the faces of people captured by the conventional CCTV cameras that are now set up in most shopping areas. The images obtained are compared with those held on a police database. In this project an alarm is set off if there is a "hit",that is when a face is recongnised. Police can then take the necessary action,for example, they can continue to monitor their target image on a screen in their own control room.

This is but another example of the application of research in image recognition producing a system that is sufficiently reliable to be used in real-life situations. Such systems will become more accurate and we can expect them to be part of any future public CCTV surveillance arrangement. The claim from Civil Rights groups that the system is but a version of "big brother is watching you" bureaucracy cannot of course be entirely ignored. The group Liberty say that "what the police call an 80 per cent success rate is what we would call a one-in-five chance of a mistake". The study of a society under surveillance will no doubt provide socio-cyberneticians with a fascinating research project.