Development and Performance evaluation of IRIS recognition in the presence of occluded region

Abstract

For secure, and reliable authentication process as compared to other security systems such as password or any other biometric systems, Iris recognition is computationally more efficient. Conventional iris segmentation methods gives good results only for ideal iris images. The segmentation accuracy has influence on the performance of non-ideal iris images. Iris recognition system comprises of Iris segmentation, Normalization, Feature extraction and Matching. IRIS databases like CASIA V1.0, MMU2, UBIRIS V1 and UBIRIS V2 are used for experimental results. Image segmentation using active contour method, the coarse parameters of iris, pupil radius will be known. The normalization process involves unwrapping the iris and converting it into the rectangular image as proposed by Daugman. The normalized iris image converted to iris code using feature extraction. Hamming distance classifier is to matching the two iris codes. The Iris images are accessed through serial communication from personal computer. The four steps are carried out for iris recognition system with the help of personal computer, controller unit (ARDUINO UNO R3), LCD display and servo motor (for activation). The IRIS recognition helps in authenticating a system.