Output details
15 - General Engineering
University of Kent
An Investigation of Quality Aspects of Noisy Colour Images for Iris Recognition
This paper examines the distributions of quality metrics in colour iris images datasets and discusses the potential impact of these variations to recognition accuracy. It introduces new metrics for the iris area affected by point-source reflections, iris diameter and pupillary boundary contrast and examines existing focus, entropy and constriction measures. There are limited published investigations which consider “colour” rather than the more commonly used infrared images. This work provides additional quantitative evidence for the suitability and limitations of colour images for iris-based biometric implementations. Funded by the European Regional Development Fund (Interreg IVA) NOmad Biometric Authentication project (ref. 4051).