Biometric Imaging Faces a Reality Check

Hype or reality: Industry insiders discuss biometric imaging and security concerns


Successful Implementations
Advanced Imaging: In what sectors or types of implementations have implementations been particularly successful?

Jim Carlson, Lockheed Martin: Transportation and security are just two sectors in which implementations have been particularly successful. Lockheed Martin has implemented systems for customers ranging from the FBI, to the Department of Defense, to the Florida Department of Education, all of which rely on accurate biometric imaging. Our recent work for Verified ID's "Clear" traveler program demonstrates the use of biometric technology in everyday life. The key to success with biometrics on all of these programs is a clear understanding of how to properly implement biometric technologies. All implementations for our customers have been successful and demonstrate the power of the biometrics tool.

Peter Kalocsai, Pelco: After many years of disappointments at various airport installations, face recognition is finally having some success stories, both domestically and abroad. You will also find today that law enforcement agencies and correctional facilities across the country are deploying biometric applications (booking systems, mobile identification systems, "spycam" surveillance systems, etc.) with considerable success. The industry is also receiving a sizable boost from the biometric national ID and passport (e-passport, digital passport) programs of ever-increasing number of countries. And biometrics is also a prime candidate tool for fighting identity theft, which is the fastest growing crime in the U.S., claiming 10 million victims yearly.

Bill Willis, ImageWare Systems: The Mexican government has selected ImageWare to develop a multi-biometric identification system that is used to help identify missing and deceased Mexican citizens. Leveraging ImageWare's Biometric Engine, the system incorporates four biometrics — face, finger, signature and DNA — that allows for a higher accuracy rate. Also, specific demographic details can also be entered such as hair, clothing, height, dental history, etc. Access control into buildings has been particularly successful, focused on specific needs and problems with great results, using the correct identity components to solve the problem.

Mike Scholten, DRS Infrared: Biometric programs have had a successful and visible history in the sectors of forensic analysis and criminal prosecutions. They have been successfully implemented in enterprise security designs when these systems occur where robust enterprise security architectures are deployed within a well-controlled user population. In these cases, success is ensured through proper countermeasure design, architecture deployment and commonsense security management operations that employ tools like user registration, certificate and credential management, and — most importantly — system attack countermeasures.

Mohamed Lazzouni, Viisage: The real success of face recognition has been in controlled environments where customers are managing large databases of images. In U.S. departments of motor vehicles, if a citizen would like to obtain a driver's license, our nation's most widely accepted identity credential, that person must pose for a photo. The DMV conditions provide a perfect scenario for comparing images in the database to find potential matches and duplicates. Trying to find identity theft and fraud in a database would have been near impossible before given the time it would take for operators to review the millions of images. Now we are able to mine these image databases, find duplicates recorded under different names and demographics, and begin to investigate suspected identity theft and fraud. The same is true for other identity credentials and large image databases around the world, including those for passports, visas and national IDs.

Are Standards On the Horizon?
Advanced Imaging: How important are standards for biometrics, in terms of acceptance and successful implementation — and what areas still need to standardize?