Iris recognition achieves highest accuracy
In 1991, a method to perform iris identification using wavelet analysis was developed by John Daugman, professor in the Computer Laboratory of Cambridge University (Cambridge, England).
In 1991, a method to perform iris identification using wavelet analysis was developed by John Daugman, professor in the Computer Laboratory of Cambridge University (Cambridge, England). To commercialize the method, Daugman formed IriScan (Marlton, NJ), a company that holds the worldwide patents on this iris-recognition concept. Today, companies such as Sensar (Moorestown, NJ), Oki America (Hackensack, NJ), NCR Corp. (Dayton, OH), and Diebold (North Canton, OH) are licensing the iris technology for the development of various people-recognition systems (see Vision Systems Design, Nov. 1999, p. 30).
Now, Spring Technologies (Falls Church, VA) and the Charlotte/Douglas International Airport (Charlotte, NC) have begun a pilot program using IriScan's iris-recognition technology to control access to restricted areas. The airport will deploy Spring Technology's TranSecure system, intended to provide substantially higher security among airport employees.
In system setup mode, digital images are taken of each employee's eye by a video camera and stored in a database. Then, during system operation, a 1-s glance by any employee into the TranSecure video camera results in recognition of the person as an airport employee. Based upon pre-established access criteria, the employee is then either allowed or denied access to the desired secure area. The duration of the pilot program is approximately six months, during which time additional phases may be introduced to increase functionality and expand the employee enrollment population. According to Spring Technologies, the performance of the iris-recognition-based system is being evaluated throughout the pilot program for comparison with existing card/password systems.
To date, Spring Technologies, IrisScan, and Sandia National Laboratories (Albuquerque, NM) have developed a comparison between iris, fingerprint, facial, and hand geometry identification methods. The accuracy/crossover error rate—the point where false-accept and false-reject rates are equal—provides a measure of the optimal system performance point of biometric methods (see table). Accumulated data prove that this error rate is orders of magnitude smaller for iris recognition systems. And, because, iris recognition techniques, such as those developed by Daugman, have never experienced a false accept/false identification, the iris error value of 0.00008% is strictly theoretical.