Eye recognition

Security systems that use the coloured part of the eye to validate a person’s identity could become more accurate thanks to new computer algorithms developed at the University of Bath.

“The algorithm has been 100% accurate in our initial trials and, even though we don’t have their exact implementation systems, it outperforms the industry standard by a long way,” boasts Professor Don Monro from the University’s Department of Electronic and Electrical Engineering.

“Accuracy is very important in iris recognition because as well as ensuring that the person in front of the technology is who they say they are, you also don’t want people turned away should the system makes a mistake; the so-called ‘insult factor’.”

The researchers now need to give the algorithm a proper work out, which is why they are in the process of expanding their database of iris images by collecting 16,000 from students and staff at the University.

The research team are also involved in two separate trials being run by UK and USA government agencies which are assessing iris recognition and other ‘living passport’ technologies for future use.

Iris recognition is seen as the most accurate biometric recognition technology because scientists regard the pattern of the iris as offering even greater discrimination between people than their fingerprints. No two irises are identical, not even between identical twins, or between left and right eyes.

Iris recognition works by using algorithms - computer procedures - to ‘unwrap’ a digital image of a person’s iris and create a unique encrypted ‘barcode’ that can be stored in a database.

When a person goes to enter a secure building, for example, a digital image of their iris can be taken, processed using the same algorithm and then compared with the barcode stored in a central database.

“There is a lot of interest in using the iris as an enhanced security measure in ID cards, airports and access to other high security areas, such as bank vaults, but the current lack of a competitive market has stifled innovation and has meant the technology is struggling to live up to its potential,” said Martin George from local company Smart Sensors who are sponsoring the project.

“A generic patent covering the use of the iris as a way of validating a person’s identity has hampered independent research and development in the field. Now the patent has lapsed, research teams around the world are looking at ways of improving the technology and building upon its obvious benefits.”

Professor Monro said: “One of the problems that we have faced in this area is the lack of a publicly-available database of iris images that we can test our software on. Most of the databases that are available are held by commercial interests, so it is difficult for independent researchers to make headway in this field. We are now making our database available online so that researchers around the world can use it to develop their own products. So far, more than 30 research groups have applied to use it.”

Professor Monro’s algorithm has just passed a major hurdle in the patenting process, with the official examination under the international Patent Co-Operation Treaty procedure establishing that their application is ‘free and clear’ of all other patents held in this area.