Indian researchers have developed a computer algorithm that can map blood vessels lying beneath the skin of the human face using thermal imaging for identification.
According to Business Standard, the new algorithm enables instantaneous recognition, which as of yet has proven to be nearly impossible to spoof. A research team at Jadavpur University in Kolkata is spearheading the initiative, and claims that the facial vessels underneath the surface of the skin are just as unique to each individual as their fingerprint or iris.
Whereas fingerprint spoofs can be forged using silicon or rubber, the researchers at Jadavpur believe it to be an almost impossible feat to create a realistic facemask for an impostor to wear that could also effectively replicate the pattern of a person’s facial blood vessels. In fact, the quality of a spoofed facemask is irrelevant as a thermal imaging camera would also pick up the imposter’s own blood vessels as well, in theory rendering the mask itself useless.
The researchers’ computer algorithm can analyze the details of the blood vessels as revealed by an infrared facial scan. The resulting thermogram maps out the pattern of blood vessels accounting for even the smallest capillary at an accuracy rate of more than 97%.
Thermograms boasting a 97% degree of precision along with a secondary identity credential – a photo ID, access card or PIN – could certainly pass for high-security applications. Add the fact that facial recognition is already widely deployed in security systems, law enforcement and legal agencies the world over, and the addition of thermal imaging begins to make sense.
For more on the research being done at Jadavpur University, see the team’s work here.