EFFECTIVENESS OF EDGE DETECTION OPERATORS ON THE PERFORMANCE OF IRIS BIOMETRIC RECOGNITION
Pillai Honey Nagarajan, Sandesh Ramesh, T. Sundararajan · 2014
Iris recognition is the most accurate biometric identification system on hand. Most iris recognition systems use algorithms developed by Daugman. The performance of iris recognition is highly depends on edge detection. Canny is the edge detectors which commonly used . The objective of the research are to a) Study the edge detection criteria and b) Measure the PSNR values in estimating the noise between the original iris feature and new iris template. In this paper Daugman's rubber sheet model is employed for iris normalization and unwrapping, descriptive statistical analysis of different edge detection operators is performed. Results show that the PSNR values of iris feature before and after the process of extraction, was 24.93 and 9.12. For Sobel and Prewitt, both give 18.5 after the process. The impact of edge detection techniques produces higher accuracy in iris recognition system.