It's no secret that some people are better at holding their booze than others. But just because some party girl appears articulate and physically composed doesn't mean she's not drunk. "Subjective analysis based on behavior is a scientifically inaccurate benchmark," says Rollin Bishop at Geekosystem. That's why researchers at the University of Patras in Greece are pioneering a new type of thermal imaging technology that can instantly detect how inebriated a person is just by scanning his or her face. Here, a quick guide to the breakthrough and who it might actually benefit:
What does it do, exactly?
While we've been intuitively guessing other peoples' level of drunkenness for centuries, computers can do it much faster and more accurately, says Michael Harper at RedOrbit. Here's the basic idea behind the thermal camera: When someone drinks, blood vessels on the surface of the face dilate. Warm blood rushes in, creating hot spots that thermal imaging cameras can detect. A similar technique has been used to screen for people infected by viruses like SARS, since a high fever is common.
How does the technology tell if someone's drunk?
The team is working on two different algorithms to analyze and assess the heat patterns on someone's face. The first counts the number of pixels the thermal picks up. Those are then measured against scans of sober and inebriated control subjects to determine if the person is drunk or not. In theory, "this should provide an accurate measurement" for drunkenness, says Geekosystem's Bishop.
What about the second algorithm?
The second compares actual temperatures readings from different regions of the subject's face. Researchers discovered that the forehead of inebriated people, strangely, shows up relatively cool in thermal scans. A person's nose, on the other hand, pops up as a hotspot. This method wouldn't compare scans to a database, but would instead compare values inside a given range to determine if someone had too much to drink.
Who would use these thermal cameras?
Because of its ability to quickly scan multiple faces at once, the technology could be helpful in all kinds of fields, from law enforcement to nightclub operation to airport security. People working at convenience stores, for example, could even use it to rapidly assess if someone is too drunk to be purchasing more alcohol at 2 a.m.