Thermal Camera Algorithms Could Scan For Drunk People In Public
Computer scientists have developed two algorithms that, if used with thermal imaging camera technology, could scan for intoxicated people in public places.
Georgia Koukiou and Vassilis Anastassopoulos of the University of Patras in Greece probably don't like alcohol and the rowdy aftermath that comes of it. In their paper that will be published in the International Journal of Electronic Security and Digital Forensics, the two different algorithims they have in the works closly analyse a subject's face. The system looks for blood-vessel dilation on the surface of your skin, a common side effect of one drink too many.
The first algorithm compares your face with a database of facial images of both drunk and sober individual, to judge your sobriety on pixel values on different areas. The second algorithm maps out the areas of the face: identifying what parts are hot and cold, basing judgement on the findings that a drunk person tends to have a warmer nose and a cooler forehead.
The paper claims that this could be a useful system for the police, so they have some more logical evidence of drunk and disorderly behaviour...not that you need it with the unfortunate state that's quite easy to identify.