Iris Recognition System using Gabor Filter & Edge Detection

Abstract

Biometric identification is the recognition of an individual by means of their physical or behavioral characteristics. Iris recognition in past decade has become one of reliable & unique biometric identification. In this research the various iris recognition method such as segmentation & normalization are applied on the eye image of 320x280 dimensions is obtained from CASIA database. 1-D Gabor Filter is used for feature extraction for eyelid detection & enhancing the segmented iris image. Edge detection techniques such as canny, sobel & prewitt are used for obtaining the fine edges. Performance is calculated on various evaluation parameters such as PSNR & MSE value for these edge detection techniques. Result shows sobel has less PSNR & showing more accurate & better result as compare to canny & prewitt. Matching is done using hamming distance by calculating the bit difference of processed or original image.