Identification and Biometrics

Fraunhofer Institute for Computer Graphics Research IGD

© Fraunhofer IGD
Competence Center Identification and Biometrics

Tabs

Outline

Not least due to the change in ID laws and the inclusion of biometric characteristics in IDs, the topic of biometrics has become increasingly important in the past years. We have been developing and working in this seminal technology field since the mid 1990's. Biometrics deals with recognition of humans by a machine: the machine draws on certain physical characteristics with which it recognizes any person already known to the machine. The most well-known methods are finger print, iris and facial recognition. Other methods include vein patterns, gait and signature recognition.

There are a great many methods and suppliers in this field and our job is to help companies, government agencies, and the general public to stay on top of this broad technology field. Our "Identification and Biometrics" competence center, which is equipped with a demonstration and evaluation laboratory, implements biometrics projects and evaluates existing biometric systems and procedures. Expert consultation, assessment and research is required because not every system is suitable for every application scenario. The field of “Identification and Biometrics” involves a broad range of activities. The expertise from our interdisciplinary team of researchers enables us to explore this topic in great depth. As an independent research facility, we are the ideal partner for users interested in biometrics.

This technology not only identifies humans using their own physical characteristics, rather similar mechanisms can also be used for multimedia files because they too possess highly individual characteristics that are a result of their content. The technology behind this is called "perceptual hashing." Significant content-related characteristics are used to determine a unique check sum by means of a one-way function. In ideal situations, similar data (images, video sequences, audio files) produce similar check sums regardless of the format used (JPG, PNG, BMP, etc.) and, ideally, also regardless of minor changes, such as the choice of an excerpt (cutting) or a change in the files size (compression), etc. For example, this allows a user to check whether already known and protected data are being exchanged in a network. We thus identify certain images, videos, or music files.

Remark: For Projects in IT-Security we base us on our collegues of the Fraunhofer SIT.

Committees & Memberships