Iris detection and extraction based on intgro-differential operator
Bo Fu, Yuanyuan Cai, Yuehao Yan · 2010
Iris, as a biometric with the best performance in the stability, reliability and non-invasiveness, can obtain an even more high recognition rate than other biometrics, such as fingerprint, face and voice. However, it is difficult to extract iris features from an eye image. In this paper, we demonstrate a scheme for iris detection and extraction based on John Daugman's Integro-differential operator. Using original image gray histogram, the center coordinates is approximately estimated in the vertical and horizontal direction, then the inside edge of the iris is determined. After that, the outer edge of the iris is computed by using the area integero-differential operator. The Gabor filter is used to extract the features of the iris and the Hamming distance between two iris features is used for recognition. The design is provably effective through the experiment finally.