A Novel Cancelable Iris Recognition Approach

A new random projection approach for cancelable iris recognition is presented in this paper. Instead of using original iris features, a masked version of the features is generated through the random projection for enhancing the iris recognition privacy. The proposed framework for iris recognition includes iris localization, sector selection of the iris to avoid the effect of eyelids and eyelashes, normalization, segmentation of the normalized iris region into halves, selection of the upper half for further reduction of the effect of eyelids and eyelashs, feature extraction with Gabor filter, and finally random projection. This framework masks the original Gabor features to increase the level of security while excluding eyelids and eyelashes’ effects. The proposed framework achieves promising recognition rates of 99.67% and an equal error rate (EER) of 0.58%.