Downloads: 108

M.Tech / M.E / PhD Thesis | Pattern Recognition | China | Volume 6 Issue 10, October 2017

Securing Iris Recognition System for Access Control Based on Image Processing

Isaack Adidas Kamanga ^[2] | Dening Jiang ^[2]

Abstract: Results of different studies show that iris recognition is a viable, fast, and easy way to authenticate a person's identity and it is the most accurate biometric approach for authentication. The findings are so positive that the prospect of iris prints replacing PINs and other identification methods is much more viable, new studies found that an eye from a dead person would deteriorate too fast to be useful, so no extra precautions have to be taken to be sure the user is a living human being. The big challenge with iris recognition is that an iris scanning security can be cheated by presenting a high resolution still image of an eye to the iris scanner instead of a live eye, also can be cheated by using video frames and 3D model of valid individual. With increasing e-payments nowadays, Un supervised passages with iris scan as a security, a hack-able security can assists criminals and others to access personal and financial information of an individual and in turn, the loss of trust, confidence and respect. This research work propose a more secured iris recognition system which incorporate the ability to detect the aliveness of the individual being scanned by the system for identification to prevent possible spoofing using photographs, videos, and 3D models of a valid user. An algorithm has been developed adapting a proposed improved canny edge detection method, this algorithm validates the user being scanned by three stages which are, Eye blink detection, Background subtraction and correlation, and Straight line border edge detection. All these three stages are performed using digital image processing techniques

Keywords: Biometric ID, Iris recognition, Hamming distance, Eye blink, IrisCode

Edition: Volume 6 Issue 10, October 2017,

Pages: 1131 - 1140

Similar Articles with Keyword 'Biometric'