Region based Segmentation of IRIS Images for Efficient Biometric Authentication using Sectional Mass Estimation Technique and Gray Level Distribution Measure

Abstract

The segmentation of iris images plays vital role in the process of biometric authentication and there are many approaches has been discussed earlier, but suffers with the problem of less authentication accuracy and segmentation efficiency. To improve the efficiency of biometric authentication using iris images, an novel sectional mass estimation technique has been proposed. The method performs preprocessing of input image and removes the noise to improve the quality of image. The enhanced image, is applied with region based segmentation using gray level distribution matrix and groups the similar pixels. From the segmented image, the method extracts the features like the size of pupil, the area occupied, and computes the gray value distribution. The same is performed for the limbic region and computes the region based distribution measure. The computed measure is converted into the feature vector and used to compute the gray level distributional sectional similarity measure (GDSSM) and based on computed GDSSM, the biometric authentication is performed. The proposed method produces efficient results in authentication and produces less false authentication ratio than other methods.