Quantum Biometric Identification

Biometric identification, the measurement of human characteristics unique for each person, is a fast growing market of $5 billion and a growth rate of 20%. This is because security has recently become a major concern globally, with an ever growing  demand for secure identification methods. 
 
A common drawback of all existing methods is that they are «classical», that is, their security is neither guaranteed nor quantified by any law of physics, but merely rests on the hope that potential impostors will not have access to the means necessary to foil them. In principle, however, they can be foiled.
 
In a recent publication at Physical Review Applied, a new journal of Americal Physical Society, Prof. Iannis Kominis of the Department of Physics of the University of Crete, in collaboration with M. Loulakis from the School of Applied Mathematical and Physical Sciences of the National Technical university of Athens, G. Blatsios from the Department of Opthalmology of the University of Innsbruck and C. S. Vrettou of the National and Kapodistrian University of Athens Medical School/Evaggelismos General Hospital, proposed a new biometric method, the security of which is based on and quantified by the laws of quantum measurements.
 
The method leads to unprecedented performance, with the probability for false positive identification lower than 10^-10. Furthermore, it is shown that in order to foil the method, an impostor must be equipped with quantum technology, which is not expected to be available for many decades, in particular quantum thermometry and magnetometry with energy resolution at the level of 10^-9  ℏ or better. 
 
The new biometric signature introduced in this paper is the optical loss light suffers along its path from the eyeball towards the retina. These losses can be estimated from the known photon number of a weak pulse of light incident on the eye and the response statistics of the tested subject on perceiving or not the light flash. Essentially, this is a complex “fingerprint” involving the eyeball, the retina and the occipital lobe of the brain. 
 
This work has the obvious potential for commercial application. For this reason, it has been commented on by MIT Technology Review. It has also been presented  in Physics World, and has been chosen by the Editors of Physical Review Applied as “Editor’s Suggestion”.

Picture: (a) Schematic of the new biometric method, in which a weak pulse of light having Ĩ photons on average is incident on the eye. Along its path towards the retina the light suffers optical losses described by a parameter α, which is modeling the transmission coefficient of an optical filter. (b) The probability to see the flash of light as a function of Ĩ , for various values of the parameter α.  This probability is based on Poisson statistics followed by the photon number of coherent radiation. (c) An identification algorithm is based on a selection and illumination of pixels on the retina having large α, so that they form a pattern recognizable to the user, and the simultaneous illumination of many low-α pixels. While the user will not perceive light in the latter, the impostor, even when equipped with an ideal photodetector, will see all pixels lighted, as in (d), and will not be able to discern any symbol.

Article: “Quantum Biometrics with Retinal Photon Counting", M. Loulakis, G. Blatsios, C. S. Vrettou and I. K. Kominis, Physical Review Applied, 8, 044012 (2017).