Biometric technology such as iris recognition uses physical characteristics -- including facial shape, fingerprints, retinal photos and iris patterns -- to confirm identities. The technology works by photographing the iris, the colored membrane that controls how much light reaches the retina, and converting the picture into a computer code. The code is compared with one in a database.
In 1936, a St. Paul, Minn., ophthalmologist named Frank Burch proposed identifying people using the furrows, ridges, rings and freckling that make every iris unique. But it wasn't until 1987 that eye doctors Leonard Flom and Aran Safir were granted a patent for the concept of the identification technology.
A Company Seeks Ubiquitous Iris Scans On PCs, ATMs and Cell Phones (PopSci.com)
Like fingerprints, every person’s irises are different; not even both irises of the same person are the same. Fingerprints can take a while to verify through state and national databases, but an iris scan, which uses more data points for biometric identification, can come up with a match within a few seconds. Those are the pros, if you’re a security expert. But they’re minuses if your concern is privacy. Fingerprints, by their very nature, are an active identification metric; you have to touch something to imprint them. But iris scans are passive — you just walk past a security camera (or, as in “Minority Report,” billboards in the mall) and the person controlling the scanner can spot your identity within seconds.The Pittsburgh Post Gazette article is straightforward and informative.
The PopSci article reads like a press release masquerading as news.
The facts the PopSci article presents seem a bit off, too.
It has always been the case that iris matching is slower than fingerprint matching. That's why it's popular in jails, where the customers aren't in much of a hurry. In places where throughput is a primary concern, such as school lunch counters, fingerprint biometrics have been more popular.
The author says that the iris match is faster because it uses more data points. That's not really the way computers work. Other things being equal, processing more data points takes more time.
Iris matching technology is highly accurate. More accurate, even, than fingerprints. In a side by side comparison between iris and finger with participating individuals where speed is not a factor, iris wins hands down.
There are many applications where iris is the preferred modality for an identity management application. But like we said yesterday, there is no magic bullet in identity management.
Iris at a distance, as discussed in the article, moves the goalposts and offers intriguing possibilities. But, especially in this industry, it is important to avoid over-promising and under-delivering.
For some background on how the company referenced in the article sees the future and their current capabilities, see the Fast Company article from last August: Iris Scanners Create the Most Secure City in the World.
