To date, fingerprint recognition is one of the more popular security features when it comes to the newer generations of the Smartphone’s as well as as biometric technology. Since the introduction of Microsoft iris recognition that is featured on their Smartphone’s, there have been some comparisons between the two biometric traits.
Fingerprint and iris recognition both feature simplicity, reliability, and accuracy when compared to the other types of biometric traits. These are the attributes that make fingerprint and iris recognition performs far better which is a promising security solution in society today.
The process begins by capturing images of fingerprint or the iris that is then pre-processed to remove noise effects. Distinguishing features are extracted which will then be matched to locate similarities between both these feature sets. Matching scores will then be generated from individual recognizers which are passed to the decision module that makes the decision on whether the person is an imposter or is genuine.
The pattern of the iris is unique to every individual and will remain constant through the lifetime of this person. The black circular disk called the pupil is positioned in the middle of an eyeball and dilates when exposed to light and will contract when exposed to darkness. An annular ring positioned between the pupil boundary and sclera is known as the iris which features a variety of details. The iris features data-rich physical structures which contain flowery patterns which are unique to each person. This particular pattern stays unchanged regardless of the person’s age and is easy to read by door locks with retinal scanners
The unique flowery patterns are what are extracted from the remainder of the captured eye-image that is transformed into a strip, and pattern-matching algorithms are then applied to this strip. The acquired image of the iris must be rich in its texture as all processes involved in iris-recognition are dependent on the quality of these images.
Today the “automated iris” recognition system has been proposed whereby multi-scale type quadrature wavelets are used to extract structure information in association to the individual’s iris. 2048 bit “iris codes” are generated whereby the difference between the representation of a “pair of Iris” will be compared in the way of comparing the Hamming distance using what is known as a XOR operator.
At the different resolution levels of the “virtual circle” of the iris image, zero-crossing representations of the 1-D wavelet transform which is what characterizes the actual texture of a retina which is calculated. The surface of the iris will be obtained using the Laplacian pyramid which is constructed from 4 distinctive resolution levels. The normalized correlation will be used to confirm whether the model image and the input belong to the very same class.
Fingerprint recognition is one of the highly proven methods used to verify the identities of individuals and has become one of the more widely utilized biometric technologies. Prints are typically made out of valleys and ridges which are found on the surfaces of the fingers. When it comes to fingerprint recognition, three main steps are used to acquire these images that use the minutiae matching approach. These steps include Image Enhancement, Minutiae Extraction, and Matching.
Fingerprint technology has been extensively used in law-enforcement communities as well as the AFIS database. It is, therefore, a highly popular type of biometric technology as well as been widely accepted. However, the fingerprint readers today are no longer able to deal with the significant variations in the populations which are required to become enrolled. Performing searches in the larger scale-deployments can take up to several minutes and may require ancillary data like sex, age, etc. for the partitioning databases to increase the search speed. Also, these searched might return some multiple matches. Therefore a back-up of the identification techniques like iris recognition will be needed to offer a resolution for the multiple matches.
Fingerprint technology is suitable for the background-check applications. While iris recognition delivers higher accuracy rates as well as being a non-invasive type of biometric technology. Iris recognition also features no false matches in more than 2 million cross-comparisons. This technology can handle far larger populations at a much higher speed