Face Recognition helps to prevent ID Fraud

Photograph search in large databases.

ID fraud has become a growing problem. For example, a person may try to get a second identity by applying for a driver’s license, ID card passport, or visa giving a wrong name. Usually, a photograph has to be delivered with the application. To find out whether the person on that photograph has already an ID document under a different name, the photograph has to be compared with all existing photographs in a database that can be as large as tens of millions of images.

Cognitec’s software product FaceVACS-DBScan assists in finding duplicate images in such large databases by exhaustive facial image analysis. Images from different sources can be quickly compared to those stored in multi-million image databases, resulting in a small list of the most similar faces that can be inspected by a human operator. The search time is in the range of seconds.

The product overcomes the problem that human inspectors cannot compare such large sets of images in a short time. Human inspectors let the product do all the inspections of non-similar identities automatically whilst controlling the similarity thresholds. They can let the system propose and present them all the similar identities in an appropriate manner. Thus, human inspectors can focus with all their concentration on the most similar identities only.

Target customers for this product include passport and National ID offices, visa issuance authorities, as well as law enforcement agencies.

Customers using this product include the Department of Motor Vehicles of the State of Alabama, The Federal Office of Administration in Germany, The Ministry of Home Affairs in Singapore, The Department of Foreign Affairs and Trade of Australia.

Summary of main features

  • Based on Cognitec’s leading FaceVACS® face recognition technology
  • Scalable and robust client-server application
  • Template density per computing node up to 2.9 million
  • Throughput up to 2.3 million comparisons per second on a standard Pentium 4 system
  • Batch and interactive enrolments and identifications
  • Enrolment image quality assessment
  • Multiple facial images per person
  • Auditable configurable journal for revision purposes and for general usage / quality reports
  • Integration Kit for easy integration into existing systems
  • Pre-binning by bin sets
  • High Availability Option
  • Runs on Windows 2000 / XP and Linux
  • Requires SQL 92 compliant database
  • Oracle 9i / 10g recommended

For more information, please contact Cognitec at:

 

www.cognitec.com

 

Top of Page


Dr. Jürgen Pampus

AUTHOR BIOGRAPHY

Dr. Jürgen Pampus,
Vice President Sales & Marketing and co-founder of Cognitec Systems

Cognitec Systems GmbH
Jürgen Pampus held various positions in the computer and security industry. At Siemens, Jürgen started up the biometrics department in 1995. This was the beginning of the development of the FaceVACS face recognition technology. This technology is now owned and marketed by Cognitec. Cognitec’s technology is used in various access control, border control, and database search solutions. One of the well-known installations has been the SmartGate system at Sydney International Airport. Jürgen is active in several biometric working groups and is co-chair of the German TeleTrusT WG6 Biometrics. Jürgen has given presentations at various international conferences on Biometrics and Airport Security. Currently, Jürgen Pampus is the Vice President Sales & Marketing at Cognitec with world-wide responsibility. This includes operations from Cognitec’s German headquarters as well as Cognitec Systems Corporation in the United States.
Telephone Number
Visit Company Web Site
View Buyers' Guide Entry

 

Print Article
View Articles in this Category
Home - Suppliers A-Z - Articles - Products and Services - News - Associations - Add Your Company - Contact Us      

Copyright 2002 Copybook Solutions LTD
All rights reserved. Reproduction in whole or in part without permission is strictly prohibited.
Site designed and hosted by .