Evaluation of Combined Visible/NIR Camera for Iris Authentication on Smartphones

Shejin Thavalengal, Petronel Bigioi, Peter Corcoran; Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2015, pp. 42-49

Iris biometrics provide a mature and robust method of authentication, but are typically applied in a controlled environment and under constrained acquisition conditions. In this paper, the adaption of iris biometrics for unconstrained, hand-held use cases such as smartphones is investigated. A prototype optics-sensor combination is analysed in terms of its optical properties and iris imaging capabilities. The corresponding camera system with dual visible/NIR sensing capabilities and 4 Megapixel resolution is tested for suitability to implement iris recognition on smartphones. Recognition performance is analysed together with image quality comparisons. Preliminary results indicate that there are challenges to achieve reliable recognition performance in unconstrained use cases. Current optical systems are not diffraction limited, particularly at NIR wavelengths; pixel resolutions are close to the useful limits for iris recognition and acquisition conditions are challenging. Nevertheless, our findings indicate a similar camera module, with an improved optics and sensor, could combine biometric authentication with more conventional front-camera functions such as the capture of selfie images.