REAL TIME IRIS RECOGNITION

Abstract

Any biometrics method refers to recognizable proof and validation of people depending upon their physical and behavioral characteristics. One of the biometric that is profoundly reliable, robust and have high recognition precision is iris recognition. However the precision of iris acknowledgment generally influenced angle, pupil dilation, limbus occlusion, intensity variations, and improper focus. So finally our aim is to increase accuracy and to accomplish high recognition rate. The different difficulties in iris acknowledgment incorporate legitimate iris region localization, segmentation, local/global feature extraction of the iris design, feature code estimate decrease, and quick calculation so that no procedure will expend much time and assets. Also, the iris recognition will work in real time.

In our project, we have set ‘HP Truevision HD' webcam with resolution of 640x480. Besides, we took the eye pictures while attempting to keep up proper settings, for example, lighting and separation to camera.

In this project, it has been showed that our algorithm gives better performance as compared to previous methods with less recognition error and 100% accuracy. Also the average recognition time taken by this algorithm is 31.01 seconds, which shows a great success of our proposed algorithm over the others.