Detailed Analysis of IRIS Recognition Performance

Koc, Oktay; Tosku, Loredana; Hoxha, Julian; Topal, Ali Osman; Ali, Maaruf; Uka, Arban

doi:10.1109/iccece46942.2019.8941784

Cited by 2 publications

(1 citation statement)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Gabor filter has the ability to generate an optimum conjoint representation of a signal in space and spatial frequency. [23] Moreover, a Gabor filter is structured by modulating a sine/cosine wave with a Gaussian.…”

Section: Encodingmentioning

confidence: 99%

Composition of Methods to Ensure Iris Liveness and Authenticity

Al-Rashid¹

2020

Adv. sci. technol. eng. syst. j.

View full text Add to dashboard Cite

In a biometric system technology, a person is authenticated based on processing the unique features of the human biometric signs. One of the well known biometric systems is iris recognition, this technique being considered as one of the most secure authentication solutions in the biometric field. However, several attacks do exist that are able to spoof iris. In this paper, we propose a novel approach for securing the iris recognition system by eye liveness detection technique. The proposed system detects the eye liveness, and recognises the iris. This process includes multiple steps. As per the first step, the person opens his eye and the system reads remotely the changes in pupil size as a result of the response to the ambient illumination. Then, the system starts matching the iris with a database. The second step: the person closes his eye and the system remotely detects the heartbeats signals under the skin of the eyelid. As per the third step, the person opens his eyes again and the system reads the pupil size again and compares the results of the pupil size, and then the system matches again the iris with the above database. For the iris recognition to be validated, all the above checks have to be passed. We have conducted several experiments with our proposed system, based on a brand new dataset comprised of 40 subjects. In addition, we also used public datasets: CASIA-Interval, CASIA-Twin and Ubiris.V1. The achieved results show the quality and viability of our proposal.

show abstract

Section: Encodingmentioning

confidence: 99%