Application of Ants Colony Algorithm for Nonideal Iris Image Segmentation

As a dependable biometric technology, identity recognition through iris has attracted attentions of researchers. Ants Colony Algorithm has been adopted to solve vast variety of optimization problems. Segmentation is of the most important parts in identity recognition system where even a tiny error is not compensable but is moved from part to part toward the end of the operation. Four optimization parameters are defined for optimal segmentation of iris using Ant Colony Algorithm (i.e. longitudinal location of center, latitudinal location of center, inner radius of iris, and outer radius of iris).The map adopted in the present study is based upon Daugman method.1877 images from 241 persons in JPEG format were obtained from UBIRIS.v1 database for analyzing nonideal iris images. Angular method was adopted in order to choose an angle of iris with lower probability of wink where overall radius (i.e. pupil to iris) is considered. Also, normalized image of Haar wavelet coefficients were used in the present study. The results obtained from the present study showed that more considerable results are achieved by using the algorithm and has got proper precision and efficiency.