Facial Recognition Technology
by: Jim Bell, August 6, 2007 5:08:00 am
"What you see here is the biometrics camera, the 3-D camera that I was referring to, that gives us the ability to get a 3-D snapshot picture of a person's face."
U of H Professor Dr. Ioannis Kakadiaris calls his process "Computational Biomedicine," which he defines loosely as a marriage of biology and computer technology. He says a flat 2-D photo is no match for his 3-D photo's high resolution detail, which is stored in the computer in the form of code.
"The computer will this 3-D information and reduce it to about a hundred numbers. Your own personal bar code. These are the unique numbers that describe your face, and make your face uniquely different from all other people."
Kakadiaris says this technology is a significant step forward from existing ID verification, such as retinal and fingerprint scans. In just one possible use, he says computer coded 3-D photos can be stored in a credit card company database. When the credit card is swiped, the store's own 3-D camera takes a photo of the customer which is compared to the 3-D photo in the database to verify that the person at the counter is the person whose name is on the card. He says putting this into widespread use could effectively stop credit card abuse and other forms of identify fraud. It can be used to control access to secure areas and facilities, and it can even keep someone other than you from accessing your computer. The facial recognition software and technology can also be used in national security. Kakadiaris sees a universe of possible uses, and he says the business world is already showing interest.
"We have a number of people that currently are interested to see the benefits of our technologies, and considering whether they want to take it to the marketplace."
As good as it is, this technology does have weaknesses. Kakadiaris says he's still figuring out what to do about natural aging, because a person's face changes as he or she gets older. He figures a new photo every few years may be the only way to solve that one. Identical twins could also fool his system, but he says solutions to these problems are just a matter of time and more research.