Rapid Development of Iris Recognition Technology

biometrics and iris recognition technology advance


In a dangerous world, being able to verify the identity of those you interact with is increasingly important. We see this in the military, in law enforcement, in civil arenas like courts and border crossings and also in commercial venues like data centers and financial institutions, where authorized-only access to information is standard practice.

The fact is, as we enter the second decade of the 21st century, virtually everyone is concerned with security to a greater degree. One could argue that this is the Age of Security.

Take, for example, our troops in Afghanistan. As they mentor Afghanis—train them as troops and to take over security functions across the country—the question lingers: Who are these individuals? Unfortunately, the inability to verify identities or confirm loyalties has resulted in multiple casualties among American and allied troops—inflicted by the same people they are training and working with, side-by-side.

In this volatile environment, Afghan soldiers must be vetted to ensure they have no criminal or insurgent links; and the key method for doing this is biometric testing, specifically fingerprints and iris scans.

 

Most secure of the secure

Biometrics is rapidly becoming the predominant method of verifying identification for a simple reason: it is more secure than other forms of identification. Of all the biometrics employed for identification purposes, iris pattern recognition is by far the most secure method.

 

Consider false acceptance rates

The false acceptance rate is the probability that an identification system incorrectly matches the biometric input pattern to a non-matching template in its database. It measures the percent of invalid inputs that are incorrectly accepted. According to the U.S. Government Accountability Office, fingerprints have a false acceptance rate of approximately eight percent. Facial recognition is about five percent. Hand geometry is two percent. Iris pattern recognition is one percent. In an increasingly tense post 9/11 world, where precise identification is vital, iris recognition is by far the most precise means to get it.

As an identifying human body component, the iris has a number of significant advantages. As an integral part of the body, it is not easy to modify. Its patterns are unique to each individual. Within the individual, right and left iris patterns are unique. The iris remains constant from the age of about a year and a half until death. So once an iris is locked into a database, the biometric is constant for life.

 

Moving forward by moving back

Within the last several years, new technology has overcome the barriers to entry that challenged the use of iris pattern recognition as a tool for verifying identification. Previously, almost all iris identification systems—many of which are still in use today—required that the individual get within a few inches of the imaging device, almost forcing the eye to touch the camera. So the user had to be highly cooperative, as this functional space violated most personal comfort zones. These systems were not easy to use.

The introduction of Iris on the Move (IOM) iris recognition systems by SRI International Sarnoff, Menlo Park, Calif., eliminates this issue; these systems are able to access iris patterns from up to three meters away and at very high throughput rates.

Portal-type versions of Iris on the Move have processing speeds of 1,800 people per hour, making them ideal for locations such as airports, security checkpoints and other applications that require uncompromising identity verification for large numbers of people. (For areas where space is limited, the technology is available in a small form factor that can still process 720 people per hour.) The technology requires limited subject interaction and is unobtrusive and eye safe. All that is required from an individual is a glance to identify his or her iris. Capture is secure through eyeglasses, contact lens and most sunglasses.

This content continues onto the next page...