Fingerprint checks and iris scans may be the most popular ways of biometric identification for the time being, but it looks like they might soon be joined – if not surpassed – by outer ear scanning.
Well, we must agree that every way has it pros and cons, and that it is highly unlikely that one will become the de facto standard for every situation, but this late arrival to the biometric party may prove to be ideal for some.
To find the idea of biometric identification through “ear recognition” suitable, one must be aware of the fact that the ear is a part of our body that changes relatively little during our life span.
“Successful ear biometrics rely upon a well enrolled dataset, with ears normalized for position, scale and rotation,” say Alastair Cummings, Mark Nixon and John Carter – three researchers from the University of Southampton’s School of Electronics and Computer Science who have set out to develop a successful technique for ear enrollment.
And according to the results presented in their paper, they have succeeded. “The presented technique achieves 99.6% success at enrollment across 252 images of the XM2VTS database, displaying a resistance to confusion with hair and spectacles,” they claim. “These results show great potential for enhancing many other already existing enrollment methods through use of the image ray transform at a preprocessing stage.”
But, the real challenge here is to find a way to use this image ray transform technique in a normal, everyday scenario, since covering part of the ear or a angling the head a bit differently when the scan is done can hinder the processing of the data. The success rate included in the paper concerns only the enrollment of the data – the first step of any biometric identification process.