Iris Location and Recognition on Mobile Devices

Hits: 0
Research Area: Uncategorized
Status: Not started  
Funded by: Fundação para a Ciência e a Tecnologia Funding: 0.00 EUR

In this project we will implement an Iris recognition algorithm on a Mobile Device, to meet the increasing security requirement and to perform a reliable authentication.

Conventionally the Iris recognition is composed by two steps, in first step, iris region is localized and in the second step, unique iris codes are extracted from the detected iris region.
Daugman method uses circular edge detection algorithm to detect the inner and outer boundaries of iris. 
In order to reduce the processing time, the pupil region can be localized in a binarized image, by changing brightness and contrast in the input image, as show in Step [1], [2].
The Iris outer boundary can now be detected by modified circular edge detection, using 10 inner and outer circle points, as show in Step [3].
Using the Cartesian to polar transformation, the normalized iris image is extracted, step [4], and the iris code generated by convolution with Gabor filter, step [5].
After that, the hamming distance was calculated between the two signatures, and the input image is classified by genuine or impostor.

[1] - Dal Ho Cho, Kang Ryoung Park, Dae Woong Rhee, “Real-Time Localization for Iris Recognition in Cellular Phone” snpd-sawn, pp. 254-259, SNPD/SAWN’05, 2005.
[2] - Dal Ho Cho, Kang Ryoung Park, Dae Woong Rhee, “A study on pupil & íris segmentation for mobile enviroment” The Fourth Berc Biometrics, Workshop, pp.99-102, 16-17, February, 2006.
[3] - John G. Daugman, “How Iris Recognition Works,” IEEE Trans. on Circuits and Systems for Video Technology, vol.14, No.1, pp21-29, Jan 2004