Alcohol is one of the leading causes of death on the road, hence the famous slogan: drink or drive, you have to choose. In Australia, researchers have developed a solution to better detect drunk drivers. As reported by BFM Tech & Co, they have designed a technology that scans images captured by “smart” video surveillance cameras.
To develop it, the researchers formed three groups corresponding to three blood alcohol levels. While participants were behind the wheel in a simulator, the technology scanned images capturing their faces. Depending on their facial features, the position of their head and the direction of their gaze, the tool was able to assess their degree of alcoholism.
75% accuracy
Quoted by our colleagues, the researchers mention an overall accuracy of 75%. This new technology would therefore be able to determine whether a person is fit to drive or not. However, they recognize the limits of this new device, which implies that the drunk driver is already on the road. He therefore already represents a danger to himself and others.
Eventually, the researchers say the technology could be used by roadside surveillance cameras or law enforcement. They just need to make sure the low-resolution images are good enough for it to work.
