Influence of cataract surgery on iris recognition: A preliminary study

Iris biometrics is considered as the unique and accurate biometric characteristics that are suited for large scale applications such as India's AADHAR, CANPASS, and many other national ID programs. However, the accuracy of the iris recognition is observed to degrade when eye (or iris) is affected by the diseases. Among many other eye diseases, cataract results in a cloud formation on the eye lens and is a potential problem, especially in the developing countries. In this paper, we present a preliminary study on 84 data subjects to reveal the effect of cataract on iris recognition performance. We investigate three different scenarios, which involve: (1) Enrolment and recognition of affected eyes (pre-operated eye) (2) Enrolement and recognition of operated eyes (post-operated eye) (3) Enrolment with affected eyes and recognition with operated eyes. Extensive experiments are carried out using five different academic and commercial iris recognition methods to get complete insight on the impact of cataract on the recognition performance. Results obtained on our database indicate the degraded performance in verification of the cataract operated eye.