Bimodal biometric identification with Palmprint and Iris traits using fractional coefficients of Walsh, Haar and Kekre transforms

Biometric identification verifies user identity by comparing an encoded value with a stored value of the concerned biometric characteristic. Multimodal person authentication system is more effective and more challenging. The fusion of multiple biometric traits helps to minimize the system error rate. The benefit of energy compaction of transforms in higher coefficients is taken here to reduce the feature vector size of image by taking fractional coefficients of transformed image. Smaller feature vector size results as less time for comparison of feature vectors resulting in faster identification. Iris and Palmprint are together taken here for bimodal biometric identification with fractional energy of Kekre, Walsh and Haar transformed Palm and Iris images. The test beds of 60 pairs of Iris and Palmprint samples of 10 persons (6 per person of iris as well asPalmprint) are used as test bed for experimentation. Experimental result that the show fractional coefficients perform better as indicatedby higher GAR values over consideration of 100% coefficients. In Walsh and haar transforms the bimodal identification of iris and Palmprint could not perform better than individual consideration of alone Palmprint but perform better than Iris. In Kekre transform bimodal with Palmprint and Iris has shown improvement in performance.