In a science known as biometrics, physical or behavioral characteristics are used for personal identification. Arun Ross, a professor at West Virginia University, explains that the sclera of the eye is another biometric trait that can be used, an idea he researched and patented. “Science of Innovation” is produced in partnership with the National Science Foundation and the United States Patent and Trademark Office.
Science of Innovation- Biometrics
ANN CURRY, reporting:
From the combat zone to the corporate office, from the airport to the ATM - knowing who's really who is a big security concern. Traditionally, a person might enter a password or pull out a driver's license or passport as proof of identity. But increasingly, identification and authentication can also require an eye scan or a well-placed hand.
CHRISTOPHER ROBISON (West Virginia University): Go ahead and turn.
CURRY: It's a science known as biometrics, recognizing individuals based on their physical or behavioral characteristics. The structure of the face, the geometrics of the hand, the ridges of a fingerprint, the patterns in an iris - every person carries multiple human traits that are a unique form of personal identification.
DR. ARUN ROSS (West Virginia University): The primary advantage of a biometric trait is that it belongs to that individual. You’re implicitly connected to it, unlike passwords or tokens or passports, which are external to an individual.
CURRY: Arun Ross is a Computer Science and Electrical Engineering professor at West Virginia University, one of three institutions where the National Science Foundation is helping to fund a coalition of biometrics research sites. The Center for Identification Technology Research, or CITeR, as the coalition is known, is an example of the important role that public and private funding can play in the innovation process.
ROSS: The National Science Foundation does offer seed funding. But that is augmented with the funding from these corporations and government agencies.
CURRY: CITeR affiliates funded Ross's 2005 research with Dr. Reza Derakhshani, who was at the University of Missouri in Kansas City. Iris recognition systems process the full iris, so when the eye's gaze is off-center, only a portion of the iris is seen, and recognition is hindered. Ross and Derakhshani proposed the idea that the sclera, the white part of the eye, might be an additional biometric trait that could be used for identification.
ROSS: We asked ourselves what type of information can we extract from the sclera in order to perhaps improve the performance of classical iris-based systems.
CURRY: First, Ross and Derakhshani collected digital images of eyes from dozens of subjects. Then they separated the sclera from the pupil and the iris to search for unique features.
ROSS: We began to see the features that were on the surface of the sclera, but also the conjunctival vasculature, the vein pattern that is evident through the sclera. So we started developing algorithms that can be used to extract features.
CURRY: Computer algorithms are step-by-step instructions that tell the computer what to do. With fingerprint recognition, they are used to extract features called minutia points, the unique points in the fingerprint ridge where it ends or splits in two. With iris recognition, they convert the unique iris patterns into a binary code, made up of zeros and ones, like a barcode of the eye.
ROSS: So when we say algorithms, really, we’re talking about image processing algorithms, algorithms that the computer program can use in order to extract features that can subsequently be used for matching purposes.
CURRY: With sclera recognition, the algorithms extract features from the sclera surface and vein patterns. As they hoped, the sclera proved to be useful in identification. Blood vessels, like fingerprints, remain stable throughout life. The computer program was able to compare and match this information.
ROSS: We began to notice that the similarity between two images of the same person was much more than between two different individuals.
CURRY: With the university's help, Ross and Derakhshani filed for a patent, which was granted by the United States Patent and Trademark Office in 2008. Simona Crihalmeanu, a research associate on Ross's team, based her own research on the patent, developing different ways to extract and match features found in the sclera.
ROSS: I think she has really contributed in taking this research to the next level by developing new algorithms for feature extraction and matching, as well as the possibility of combining the iris with this sclera.
CURRY: The research also caught the attention of a commercial venture, EyeVerify, which has licensed the patent for mobile applications.
ROSS: We have to keep moving forward, looking at new ways to solve old problems. Because, what was unsolvable a few decades ago might be solvable now.
CURRY: Solving old problems takes innovative thinking, and because of dedicated funding, researchers like Arun Ross have been able to work on new solutions - with an eye toward the future.
WASHINGTON — Do you know where your student is? At school? On the bus? Paying for lunch in the cafeteria?
Principals in thousands of the nation’s schools know the answer because radio frequency chips are embedded in students’ ID cards, or their schools are equipped with biometric scanners that can identify portions of a student’s fingerprint, the iris of an eye or a vein in a palm.
Science of Innovation, Biometrics, Technology, Engineering, Arun Ross, West Virginia University, WVU, Identification, Sclera, Human, Trait, Characteristics, Research, Patent, Reza Derakhshani, Security, Proof of Identity, Authentication, Facial Scan, Hand Geometry, Fingerprints, Eye, Iris, Scan, Recognition, Center for Identification Technology Research, CITeR, Computer, Program, Algorithms, Minutia Points, Binary Code, Veins, Vascular, Blood, Vessels, Simona Crihalmeanu, EyeVerify, Innovation, Funding, United States Patent and Trademark Office, USPTO, National Science Foundation, NSF