Clemson Facial Recognition Research Advances

CLEMSON, SC – August 12, 2009 – It often takes a pristine look at the iris to pass through some security systems. Today with the help of the Clemson University Image and Video Analysis Lab, systems may just need a wrinkle to verify identity.

The lab is working with a $2 million grant from the Office of the Director of National Intelligence and has joined with other universities to create the Center of Academic Studies in the Identification Sciences (CASIS). The aim of the center is to strengthen biometric identification, the measurement of physical characteristics to confirm a person’s identity.

This image shows some areas of the eye that can be used for identification.

photo by: Damon Woodard

At Clemson, School of Computing assistant professor Damon Woodard has led the way in developing periocular- and iris-recognition techniques, which can achieve high performance when using less-than-ideal quality data.

“Historically, we have looked at the iris for identification, but what if the iris is obstructed when the subject blinks? We then can look at eye region features such as skin folds and wrinkles for more accurate identification along with the iris,” said Woodard. “Zeroing in on these features allows for the identification of individuals when a large portion of the face is obstructed or unclear. When we can combine the iris with periocular features, the result is a more accurate system.”

Woodard believes the techniques could open up additional possibilities for applications that require the identification of an individual, including airport, border and home security.

Under the five-year grant, which totals $8.9 million, Clemson is working with researchers from Carnegie Mellon University, North Carolina A&T State University and the University of North Carolina at Wilmington.