IRIS RECOGNITION USING DISCRETE WAVELET TRANSFORM

Iris recognition is known as an herently reliable technique for human identification. In this paper, a DWT based quality measure for iris images is proposed. It includes preprocessing, feature extraction and recognition. The preprocessing steps such as conversion of color to gray image, histogram equalization and segmentation are carried out to enhance the quality image. The area of interest is selected from which the features are extracted and polar to rectangular conversion is applied. Features are extracted using DWT the templates are generated and match with the stored one using hamming distance and the FAR and FRR is calculated. The algorithm is tested on iris images of UPOL and CASIA. The accuracy of algorithm is found to be equal to 100 % while FAR and FRR is equal to 0 % as it donot accept the image which is not present in the database (unauthorized person) and donot reject the authorized person. The DWT is consider for recognition process as it is less affected by pupil dilation and illumination along with this it works better in noisy conditions. KEYWORDS:Biometrics, DWT, hamming distance, histogram, iris recognition, segmentation. I