Iris Acquisition on Smartphones (2013-2016)

Iris recognition is the method of recognizing individuals based on their iris pattern. Iris of the human eye is the annular region between the pupil and sclera, and controls the amount of incident light on retina.
Unlike the PCs and laptops, the smartphone is a very personal device, where users store their personal data, passwords, health information, financial transactions, daily calendars and to do list and much more. Hence, secure authentication of a smartphone user’s identity is crucial. The objective of this work is to provide a practical iris recognition solution for user authentication in next generation smartphones.

For smartphones, a typical use case is the user authentication while holding the device at a comfortable arm’s length (approximately 15-30cm from face) as in Figure 1.



Figure 1: A typical smartphone user authentication scenario. The distance between camera and eye is approximately 250mm.

Design Considerations:
The key element for acquiring high quality iris images is the imaging system. A dedicated iris acquisition system is needed to provide a working solution based on the technology available in today’s smartphones. The key design question is whether to use two different front facing cameras (one for general purpose applications such as video call and a second one dedicated for iris recognition) or a hybrid camera which can be used for both iris recognition and other general use cases. These options are depicted in Figure 2.



Figure 2: Iris acquisition on smartphones. (a) Dual camera system – one camera for general purpose use such as video call and one dedicated iris camera for iris acquisition, (b) hybrid front facing camera – This camera will acquire both RGB and NIR images simultaneously.

The primary design consideration can be summarized as in Figure 3.



Figure 3: Summary of primary design considerations for smartphone iris acquisition device. Our recommended choice based on the performance and cost-effectiveness is shown in the highlighted selection..

Proposed Solution: Dual Function RGB/NIR Camera for Smartphone Iris Recognition
Our proof-of-concept hybrid RGB/NIR camera for smartphone iris recognition is shown in Figure 4



Figure 4: RGB/NIR hybrid camera for smartphones. (a) Prototype device. (b) Smartphone form factor optics. (c) Hybrid sensor used in the prototype device – R, G, B and IR represents Red, Green, Blue and Infra-Red sampling filters respectively.

Iris Liveness Detection for the hybrid RGB/NIR camera:
As the use of iris biometrics on smartphone devices becomes more widely adopted, it is to be expected that there will be similar efforts in the research community to beat the biometric by exploring new spoofing methods. Hence, simple, cost-effective yet powerful liveness detection should be incorporated on smartphones. Such a technique ideally should not require any additional hardware or user interaction and should be computationally light enough to be embedded in the smartphone camera pipeline or dedicated iris recognition digital signal processors.
Such a solution is presented in Figure 5.


Figure 5. Workflow of the proposed liveness detection technique.

Iris Pattern Obfuscation:
Iris obfuscation in digital images can be a potential solution to some of the privacy concerns which may arise with the wider adoption of iris recognition. Iris pattern obfuscation aims to detect the iris region in the images and replace it in a way that the replaced images will look natural and similar to the subject’s original eye, but the iris information should not match to that of the original iris. This prevent the possibility of using iris images from a personal photograph for spoofing attacks on iris recognition systems.
Examples of iris pattern obfuscation is shown in Figure 6.



Figure 6. Examples of iris pattern obfuscation – (a) Original image (from UBIRIS database), (b) and (c) obfuscated images using different techniques. It can be noted that obfuscated images ( (b) and (c)) maintain same colour and appearance as the original one ((a)).

Further Reading:
A detailed overview can be found here:
The full description can be found here: