Biometrics researchers show iris recognition accuracy reduced by type II diabetes
The biometric accuracy of Iris recognition is reduced by the presence of type II diabetes, according to an academic study reported by Science Discoveries.
Eye conditions such as retinopathy, cataracts, and glaucoma can be caused by diabetes, and a study published in Biomedical Engineering investigated the impact of the medical condition on the accuracy of iris recognition. The researchers note that previous studies have indicated the possibility that false rejection rate in iris recognition may be significantly influenced by factors including ethnicity, gender, and eye color.
An Iri Shield USB MK 2120U device was used for biometric capture, and images compliant with ISO/IEC 19794-6 standard were used for testing.
From a database of more than 1,300 iris images, researchers Mohammadreza Azimi, Seyed Ahmad Rasoulinejad and Adrzej Pacut used four different matchers, three open source and one closed commercial offering, and found that accuracy was consistently higher with those who do not have diabetes.
Only irides without visible impairments were used in the study, and the researchers found that iris disorders are often not obvious. They also concluded that differences in pupil dilation between different age groups should also be considered a possible source of error.