Multimodal Eye Biometric System for Unconstrained Eye

Biometrics frameworks have essentially enhanced individual authentication, playing asignificant part in personal, national, and global security. Existing ocular biometricsystem achieves good accuracy results for images acquired using NIR cameras in idealcondition only. When visible wavelength images are acquired in unconstrainedenvironment, noise is introduced such as illumination, reflection, motion blur etc.which degrade the recognition performance. This research presents a multimodal eyebiometric framework utilizing Support value-based fusion (SVBF) matching process toenhance biometric authentication by combining the of iris, sclera, and pupilcharacteristics from unconstrained coloured eye images. A multimodal biometricarchitecture using the fusion-associating support-value method is introduced in thisreport to improve biometric authentication. The proposed strategy is portrayed insubsequent steps; initially CNN (Convolutions Neural Network) segmentation basedon quality feature selection using entropy is applied to cluster iris, pupils and scleraregion. Subsequently effective features are extracted from the segmented iris, pupil andsclera region, for example colour histogram, Log Gabor and sclera Y- shape features.On the basis of the extricated features, the support value-based fusion is determined,and the matching score is calculated by means of the minimum, maximum value andsupport value derived from the features. Finally, authentic person is predictable bycomputing a Euclidean distance of training and testing matching scores. The proposedfindings are tested on constrained database MMU, unconstrained image databaseUBIRIS.V2 and mobile image database MICHE to show with the current techniquesthe efficiency of the proposed authentication technique. Experimental results showsthat proposed multimodal biometric system provides better results as compared toexisting state-of-art. Segmentation performed using E-CNN improves results forsegmentation accuracy up to 97.99% for iris, 98.08% for sclera and 99.43% for pupilsegmentation under uncontrolled environment by reducing segmentation time up to0.9sec. Proposed SVBF framework also highlights the role of feature level fusion toenhance the recognition accuracy up to 97% for unconstrained visible wavelengthimages.