25.01.2019 change 25.01.2019

Identity theft will be an impossible mission - thanks to Polish scientists

Photo: the Military University of Technology press materials Photo: the Military University of Technology press materials

Printed silicone masks are one of many methods of identity theft. Cameras at border crossings might fail to identify a person impersonating someone else - unless the identity thief is scanned by a biometric system co-developed by scientists from the Military University of Technology.

"We are working on introducing a new border check method using biometric systems. We are currently at the stage of integrating all subsystems of the project and putting it together" - says Dr. Marcin Kowalski from the Security Systems Team of the Institute of Optoelectronics, Military University of Technology. According to the Military University of Technology spokeswoman Ewa Jankiewicz, 10 consortium members from the European Union Member States participate in PROTECT (Pervasive and UseR Focused BiomeTrics BordEr ProjeCT). Polish researchers are responsible for the development of the method of face recognition in the far infrared.

"Based on the thermogram from the camera, we conduct facial verification. This method has various interesting properties, but so far in biometry it has been known mainly in scientific research" - adds Dr. M. Kowalski.

In his opinion, thermographic camera is a great complement to the cameras operating in the visible light range. The developed method allows to very accurately detect identity theft attempts, such as silicone masks that are difficult to detect with ordinary cameras. In addition, the method works in difficult lighting conditions, which is especially important in the case of land border crossings.

IDENTITY VERIFICATION IN MOTION

Currently, identity verification is carried out in a static manner. The person subject to border control stands in front of a device called the biometric gate. There, the person`s face and fingerprints are scanned. "In the system we are working on, we have changed the rules of verification so that it can be carried out in motion. Persons subject to identification do not have to stop. We have developed a biometric corridor, in which various sensors are located (face, retina, gait sensors). These sensors, colloquially speaking, scan a person going through this corridor. Verification takes place dynamically" - explains Dr. M. Kowalski.

Researchers are working on determining the characteristic features of people, that is, biometric modalities. Systems analysing these features have not yet been used for travellers` control. "Existing systems usually identify people based on a photograph of the face, retina and fingerprint. As part of the project, individual research centres are working on facial recognition in the infrared range, gait verification and analysis of the of the hand and finger veins image. Thanks to the use of several biometric modalities, this system will allow to achieve a much greater identification accuracy" - says Dr. Marcin Kowalski.

PASSPORT EXCHANGE WILL BE NECESSARY

The system will require passports exchange, because the ones currently in use cannot store so much data. Documents with special chips are already being developed by one of the consortium partners, the German company Veridos.

Airport is a natural environment for testing the capabilities of the system, but - according to the university spokeswoman - other border crossings can also take full advantage of the new system`s capabilities. The project will end with two demonstrations of technologies that will take place in June in two different locations, at a land border crossing in Poland and at a railway crossing in the UK.

The budget for the project implemented under the EU`s Horizon 2020 programme is EUR 5 million. The research began on September 1, 2016 and will end on August 31, 2019. Prototype biometrical corridor equipped with the PROTECT systems was presented at the Security Research Event in Brussels in December 2018.

PAP - Science in Poland

kol/ agt/ kap/

tr. RL

Copyright © Foundation PAP 2019