Date of Publication :30th March 2018

Abstract: Now a day’s iris patterns play a vital role in gender classification. Iris patterns are distinctive and cannot be changed until it becomes unnatural. This paper focus on gender has been detected using iris images. This work presents a new method for gender classification based on features of the iris texture selected by mutual information to improve gender classification of iris images For determining the recognition performance of the system database of digitized grayscale eye images were used. This grayscale images are segmented. Traditional iris segmentation methods always time-consuming and sensitive to noise. Here, in iris recognition, the segmentation is based on Hough transform used for automatic segmentation and able to localize the circular iris and pupil region. Edge points of iris boundaries are detected with canny edges steps and the threshold values are matched with the hamming distance. The Hamming distance was employed for classification of iris images. This work comes to the conclusion that iris segmentation is an essential part of recognition system and the prediction is based on iris texture features and stop matching when a generation of Iris close match is found

Reference :