This item has been filed in | Features, News
Print This Post Print This Post

Strengthening security by never forgetting a face

College of Engineering professor Mohamed Abdel-Mottaleb’s research in facial recognition technology could help tighten worldwide security

Mohamed Abdel-Mottaleb, center, and research assistant and doctoral student Steven Cadavid, right, test the facial recognition system.

Mohamed Abdel-Mottaleb, center, and research assistant and doctoral student Steven Cadavid, right, test the facial recognition system.

A drug trafficker attempts to enter the United States upon arriving at Miami International Airport. But law enforcement officials quickly swoop in on the suspect and make an arrest, thanks to a state-of-the-art scanner that photographs him and compares his image to a database of known criminals.

Such is the stuff of Hollywood. But soon, the flawless use of biometrics—identifying people by unique characteristics such as irises and facial features—will move beyond the realm of popular culture, depicted in films such as Gattaca, and into reality.

Just consider: A number of U.S. states now use facial recognition technology when issuing driver’s licenses. Even the U.S. government employs a similar method for verifying the identities of international travelers holding non-U.S. passports.

Historically, this type of technology has been expensive—especially 3-D facial recognition—and somewhat unreliable.

But now, University of Miami College of Engineering professor Mohamed Abdel-Mottaleb and his research team have developed ways to make biometric modeling more accurate while reducing its cost.

“The most straightforward approach to doing 3-D face recognition is to use 3-D capture devices like laser range scanners, which are costly and don’t provide accurate measurements of the nose, eyes, and other distinguishable landmarks,” explains Abdel-Mottaleb, professor and chair of the college’s Department of Electrical and Computer Engineering. “We did it in a more economical way, by using off-the-shelf cameras.”

He used those cameras to take images of test subjects and then employed a “stereo-based technique” that automatically locates the distinguishable landmarks of each face and calculates the depth for every point on the face, including the eyes, the nose, the mouth, and the other peaks and valleys that make up one’s facial features.

“No single approach can give you 100 percent accuracy,” Abdel-Mottaleb says. “One way to increase the accuracy is to use different biometrics and then combine them.”

Under lab conditions, his multinodal 3-D facial recognition technique yielded a 99 percent accuracy rate.

In a study funded by the U.S. Department of Homeland Security, Abdel-Mottaleb collaborated with researchers at the FBI and at West Virginia University, developing a better method for identifying people from only an image of their ear, a much more difficult identification process given the ear’s lack of readily distinguishable characteristics.

To accomplish this, Abdel-Mottaleb used video sequencing, taking panoramic images of test subjects’ heads and by building 3-D ear models. The models were then compared to others from a database of 1,000 known subjects; again, showing a 95 percent rate of accuracy.

Abdel-Mottaleb’s third biometric profiling project, funded by the National Science Foundation and the U.S. Department of Justice, aims to automate the forensic identification of subjects by dental records, significantly reducing the amount of time forensic odontologists now spend authenticating unidentified persons.

Not only could his high-tech identification tools help win the war on terror, fight crime, and enforce border security, they may also offer various household applications. For example, when a person arrives home from work, the device could scan the resident’s face then respond by turning on the lights, the television, or both—integrating with newly advancing “Smart House” technologies.

Abdel-Mottaleb’s work could even prove useful in understanding more about developmental disorders of the brain. Partnering with Daniel Messinger, UM associate professor of psychology, they are developing automated measurements of facial expressivity and gaze direction in infants at risk for autism. Measurements of these dimensions are identifying factors associated with the development of autism symptoms.

Abdel-Mottaleb describes his research as “satisfying, especially when you know that what you’re doing has real-world applications that will benefit people and enhance personal security.”

Share this Listing:
  • Twitter
  • Facebook
  • MySpace
  • LinkedIn
  • Google Bookmarks
  • NewsVine
  • Digg

Comments are closed.

  • Related Stories
  • Tags
  • Popular
  • Subscribe
  • Subscribe to the Veritas RSS Feed
    Get updates to all of the latest Veritas posts by clicking the logo at the right.

    You can also subscribe to specific categories by browsing to a particular section on our site and clicking the RSS icon below each section's header.