Award Date

2009

Degree Type

Thesis

Degree Name

Master of Science in Electrical Engineering (MSEE)

Department

Electrical and Computer Engineering

Advisor 1

Shahram Latifi, Committee Chair

First Committee Member

Mei Yang

Second Committee Member

Venkatesan Muthukumar

Graduate Faculty Representative

Kazem Taghva

Number of Pages

56

Abstract

Using Daugman's algorithm and comparable alternatives, we find that we are able to identify an iris with as little as less than half of the iris information available, and an equal error rate comparable with that of popular biometrics like the fingerprint and face recognition biometrics. Different experiments were done based on percentage of iris shown, the resolution of the iris, and the position of the iris covered to determine if partial iris recognition is a viable biometric. It was found after over 500,000 different iris comparisons amongst five different experiments that regardless of the model used and the resolution, the equal error rates of partial iris recognition were competitive with its more popular counterparts. There is a slight decrease in the equal error rate in partial iris recognition, but not nearly as drastic as expected.

Keywords

Biometrics; Iris recognition; Partial; Security

Disciplines

Electrical and Computer Engineering

Language

English


Share

COinS