Output details
15 - General Engineering
University of Kent
A Colour Iris Recognition System Employing Multiple Classifier Techniques
Iris recognition typically requires high-quality near-infrared iris images. For mobile device applications it is desirable to operate with visible light, maintaining performance. We designed a system to cope with colour iris images. Results show higher performance than existing systems. Our main innovation is using Multiple Classifier Systems to fuse diverse information extracted from colour channels enhancing recognition accuracy. This journal paper was an invited submission following feedback from the organizing committee of the 9th International Conference on Signal Processing, Pattern Recognition and Applications, 2012. The work was funded by the international ERDF Interreg IVA NOmad Biometric Authentication project (ref. 4051).