Incredible Iris Identifiers
There are iris imagers today that, whilst very efficient, may soon be right out of date, the latest much more sophisticated in being able to verify not only your identity but also your race and gender.
There are iris imagers today that, whilst very efficient, may soon be right out of date, the latest much more sophisticated in being able to verify not only your identity but also your race and gender. Today’s, commercial recognition systems use a John Daugman algorithm developed at Cambridge University in 1992 and patented worldwide.
He had computerized a process mathematically analysing random patterns in the iris to create the binary iris code, so individual to a person that only a 70% match is needed for successful identification, because the possibility of a greater than 70% match between any two irises are less than ten billion to one.
The iris is that part of the eye which gives a person’s eyes their colour and controls pupil size, growing into a uniquely complicated pattern during foetal development, remaining the same throughout life, a fact successfully exploited in iris-based biometric systems, on the principle that no two irises are ever alike.
University of Notre Dame professor Kevin and colleagues have now developed a system able to detect similarities between irises, initial tests demonstrating that the new system can distinguish between different racial backgrounds and possibly also determine gender ot the subject scanned.
Typical iris scans involve a camera capturing an image of the eye bathed in near-infrared light. Software analyses 1024 sample regions of the iris, seeking patterns in the way the stroma reflect this light, that unique information generating a binary number code.
With the slightly more complex new technique software identifies other stroma features like lines, spots and brightness variations, saving the information, in a richer set of attributes which allowed researchers to look for common features of known ethnicity and gender.
Turning the system to address a database of 1200 unknown irises, it correctly identified gender 62 per cent of the time, a rate the research team claim will improve dramatically once they have fully worked out which textural features of the iris correspond to gender.
Apart from the obvious edge this will give when testing people attempting to fabricate a false race or gender identity, the new scans could speed up large iris databases searches, making it possible to count the number different ethnic backgrounds entering a country without recording identities.
Since iris patterns are so highly random that even a person’s left and right irises differ, anything that can narrow down the possibility of identity fraud has to be considered a good thing, in this age of heightened awareness of possible terrorism threats, so the sooner this technique is brought on line the better.