Fingerprints have long been used by government agencies to identify people. The United States FBI database alone holds 70 million fingerprints of criminals. But fingerprinting isn’t an exact science. Fingertips can be damaged with calluses, and fingerprints can rub off over time.
Iris scanning is also difficult, since the eye is such a small target and subject to fraud – it is even possible to use a hi-res image of an eye, rather than an actual eyeball, to fool the machine.
With these disadvantages in mind, scientists at the University of Southampton worked to develop a new technology that would be able to identify someone via a part of the body that barely ever changes: the ear. “When you’re born your ear is fully formed. The lobe descends a little, but overall it stays the same. It’s a great way to identify people,” Mark Nixon, the leader of the research, told Wired Magazine.
How it works
According to Wired Magazine, ear identification uses computer vision to convert human features into IDs. Researchers in Southampton used this technology to create an “image-ray-transform” algorithm that can identify an ear with 99.6 percent accuracy.
The ray-producing algorithm looks for curved features in the ear. After the rays have found every part of the ear, another program turns these curves into a unique set of numbers – basically an “ear ID.”
Ups and downs
Like all technology, there are positives and negatives to using ears to identify people. Like iris scans, ear identification systems have limitations like hair covering the ears, difficult lighting conditions, and different IDs generated by different angles. For this reason, researchers stress that using the ear isn’t about replacing existing biometrics like fingerprints – but rather supplementing those technologies with another ID system.
Research into ear identification systems is still in its infancy, but as it gets quicker or more accurate, it could be used in many situations – like when a security camera grabs a profile view of a man robbing a bank. As Nixon told Wired Magazine: “We’ve shown we can use ears, but can we process the data that comes from a sort of normal scenario? That’s the real challenge.”