Keeping an eye on biometrics

Citation

Loughlin, C. (2006), "Keeping an eye on biometrics", Sensor Review, Vol. 26 No. 1. https://doi.org/10.1108/sr.2006.08726aaa.001

Keeping an eye on biometrics

Our theme for this issue is biometrics which principally concerns the identification and recognition of people; however, as our “Viewpoint” recounts it is also the term used for the mathematical analysis of biological data. We also have a sub-theme of “complex systems”. These three areas of study are currently only loosely connected however there is considerable scope for cross-fertilisation of ideas and techniques.

One of the first biometric fields was signature recognition and I can remember trying out a state-of-the-art system over 30 years ago. That system was not based on a static image comparison between a sample and a stored master but instead was a dynamic system that analysed not just the shape of each curve and swirl but also the time that was taken to draw each element. A sort of 3D analysis of a 2D image with a third dimension of time thrown in for good measure. Quite fancy stuff, especially considering that it had to be tolerant of considerable bona fide variation, while detecting visually identical forgeries. In essence this system was also measuring the characteristics and performance of the person's hand and lower arm as well as a strong element of “mood”.

I do not know what happened to the system I saw then as I have not come across it since, but here in the UK at least we have a strong trend away from signatures towards four digit PIN numbers. Mainly because earlier systems relied on a hurried salesperson to check scrawled signatures and I imagine it would have been easy to forge something that passed this cursory examination.

People are incredibly good at recognising people and I would hazard a guess that each of us is able to recognise several thousand different people and that we could identify them quite readily from within a crowd. It is therefore somewhat ironic that we are replacing a biometric such as signature recognition with a four digit PIN code, especially when each of us has enough difficulty remembering our own code let alone someone else's.

Identity theft and acts of terrorism are the main driving force behind biometric technology. There is little doubt that efforts in this area will only increase and that the size of the biometrics industry will grow accordingly.

On balance I welcome the insinuation of biometrics into our society but not without some caution. Provided that the system is secure and that I knowingly participate in each biometric acquisition then I am relieved both that my credit card account is protected and that I am spared the task of remembering my PINs. The main difficulty comes in making the system easy to use while at the same time making it impossible (or at least very difficult) for someone to steal my biometric identity.

For example, if someone could extract an iris scan from me without me knowing then they have stolen my identity as effectively as if they had intercepted a PIN code, but with the added danger that I cannot change my iris whereas I could quickly choose an alternative PIN. Of course it can be argued that it is impossible to fake an iris and so no one could copy it as a physical unit. However, at some point every iris scan is turned into a computer file which, as we all know, is child's play to copy and duplicate.

Clive Loughlin