Real-time tech to nab DUI road offenders

Police in Jilin province plan to use facial recognition technology to identify and catch drunken drivers by the end of this year, according to local traffic police.

Experts will help set up a physiology-based model containing facial features commonly identified with drunken drivers, such as enlarged pupils, flushed cheeks, shaky breathing and yawning.

By scanning the faces of drivers and comparing them with the model, high-definition surveillance cameras will be able to spot drunken drivers and send data to police in real time, including vehicle license plate numbers, so that they can locate the suspects and conduct breath tests.