Partial Template of Human Iris Patterns Recognition and identification using Neural Networks

Abstract: Iris recognition is one of the most widely used biometric techniques for personal identification. Of all biometrics- based techniques, the iris-pattern-based systems have recently shown very high accuracies in verifying an individual’s identity. Each human iris has its unique visual pattern and local image features also vary from region to region, which leads to significant differences in robustness and distinctiveness among the feature codes derived from different iris regions. However, most state-of- the-art iris recognition methods use a uniform template size (20 X 480 ) , where template generated from different patterns of the same person. In this paper the Iris recognition has been carried out employing a template of size 20 X 480 pixels , 10 X 480 , and 10 x 360 pixels. The results of the sizes of the templates have been compared and it has been observed that the accuracy of the results obtained with the limited template size is comparable with that of the one with the full size. The reason is the reduction of the space requirement as well as time complexity with no loss in accuracy. The results of iris recognition performed applying Hamming distance, Feed forward back propagation, Cascade forward back propagation, Elman forward back propagation and perceptron as presented in this paper. It has been established that the method suggested applying Cascade forward back propagation provides the best accuracy in respect of iris recognition with no major additional computational complexity.

Keywords: Iris recognition, Biometric identification, Pattern recognition, Automatic segmentation