Ophthalmologists were the first to recognize the fact that
human iris patterns can be used for personal identification [1].
John Daugman [2] who devised an algorithm exploiting
integro-differential operators and Gabor filters, developed the
first iris recognition software. The inner and outer boundaries
of the iris were detected by the differential operators and Gabor
filters were used to extract unique binary vectors constituting
the iriscode from the local texture phase information [3, 4].
The average Hamming distance between two codes was
considered for matching. Many other techniques as referred in
[5] were developed that attained similar performances. Most of
these techniques, consider single metric or criterion to identify
an authentic person from an impostor. In [5] new and simple
novel techniques, based on multidimensional multiple metrics
were proposed based on SVD, DCT, and wavelet transforms.
A full size rectangular iris code obtained following [2] is
mostly used for identification. In [6] experiments with partial
iris codes as well as compressed partial iris codes are also
shown to give very good performance in the recognition
system.