Researcher Developing Biometric Technique To Identify Suspects Seeking To Avoid Detection
Identifying a terrorist traveling incognito among passengers in a crowded, busy airport can be a security challenge akin to looking for the proverbial needle in a haystack.
An Indiana University-Purdue University Indianapolis (IUPUI) professor has received a $300,000 military grant to develop a video surveillance system for homeland security that uses a biometrics technique -- iris recognition -- to identify suspects seeking to avoid detection.
Yingzi (Eliza) Du, assistant professor of electrical and computer engineering at the Purdue School of Engineering and Technology at IUPUI is one of 33 recipients of the prestigious 2007 Office of Naval Research Young Investigators award.
Under a three-year Young Investigators award project titled “Selective Feature Based Iris Recognition for Non-cooperative User Identification,” Du will research and design software that would make it possible to monitor and identify terrorists and other criminals covertly in real time using the patterns of the irises of their eyes.
Such iris recognition “provides a new means for surveillance and terrorist watch. It is expected to have a significant impact on the military, homeland security, and intelligence, such as border control, monitoring insurgent/terrorist/criminal activities, and remotely identifying people,” said Du, whose research expertise areas include biometrics, digital image processing, pattern recognition, and their applications.
The use of biometrics -- fingerprints, face patterns, and eye or iris patterns -- is becoming more convenient and secure compared to traditional methods of identification and verification imperative to security, intelligence, law enforcement and e-commerce.
Because the patterns of each of a person’s irises are unique, iris recognition is the most accurate and reliable of form of popular biometrics identification.
“However, there is no iris recognition system that can perform positive human identification in video surveillance,” Du said. In addition, the challenge is to identify a suspect who may be facing away from the camera because off-angle iris images are often captured out of focus or with motion blur.
Du’s proposed system will automatically select iris patterns with sufficient quality to recognition. The captured patterns will be compared to those on file in a database of known subjects.