Overcoming the challenges of iris scanning to identify minors (1–4?years) in the real-world setting

摘要

Abstract ObjectiveBiometric identification techniques for pediatric use are limited. This investigation studied iris scanning in minors aged 1–4 in two exploratory studies in Belgium (n?=?197) and Sierra Leone (n?=?230), and in a subsequent clinical study in Sierra Leone (n?=?635). Images of participants’ irises were captured using a camera, while a survey assessed the ease of use with children.ResultsThe image capture success rate per individual was high; 86.0% of the participants had ≥?2 successful captures. Iris scan quality and surface were similar in all age groups and in the matching population database. When including feasibility in the analysis of minors aged 3–4, sensitivity and specificity were non-inferior compared to using the biometric of a guardian. However, the quality of iris scanning in minors aged 1–4 was worse than the iris scanning reference quality in adults. A mean total usability score of 1.55?±?0.27 was calculated; a usability threshold of 1.45 is required for routine use. Overall, this technique is feasible in minors aged 3–4, replacing the use of guardian biometrics. Additional work is ongoing to improve this technique further, striving for uniformity from the age of 1.