Dtect’s infra-red people and vehicle counter
July 31, 2007 Traffic jams, aside from being extremely frustrating, are also extremely costly. The Texas Transport Institute calculated that during 1996 traffic congestion ate up $74 billion in lost hours and petrol – and with congestion only rising since then it’s a sure bet that even more money has gone down the drain while commuters get nowhere. In an effort to combat congestion, many governments are looking towards car-pooling, offering monetary incentives to businesses that encourage it and introducing high occupancy vehicle lanes. However, a potential problem lies in policing them. To that end, Loughborough University has released the Dtect system – an infra red camera and image processing unit that can determine how many occupants a car has in an instant. And unlike a sleepy toll booth operator, or other photography based systems, it won’t be fooled by dressed up mannequins.
The Dtect system operates by projecting two wavelengths of low intensity infra red light at the oncoming vehicle. The device has a range of up to 50 metres, and is effective on cars traveling at up to 80 miles per hour, eliminating the need for them to stop or even slow down. As the beams are fired, two digital cameras, specifically coordinated to capture the infra red wavelengths, take a photo. The accompanying software combines the two images and eliminates non-facial aspects of the photo before logging the picture with a printed timestamp, location and person count. An instant after the process begins and the beams are fired, the final picture will be sent off to a station of the operator’s choice – with the faces masked to prevent an invasion of privacy. When the Dtect system is correctly set up, and has an unhindered line of access to the windshields of oncoming automobiles, it is correct in its occupancy count 90% of the time. And in the ten percent of cases when it is mistaken, and challenged, the image in question can easily be examined by a human eye.
Although the Dtect system is most likely useful in a tolling scenario and to encourage a fairer system than the flat congestion charges currently in place in several major cities, possible applications for the technology extend well beyond this. Dtect could be used as a tool for border security, and also as a data gathering tool for governments or corporations. The latter application could prove especially useful in providing the basis for better planning and strategies to combat traffic congestion and help to stem the flow of all that wasted time and money.