Comparative Analysis on Different Wavelet Families to Improvise Iris Recognition System

It is required to obtain the most discriminating data from the iris pattern in order to successfully identify a person. The validation of feature extraction, a key component of iris recognition, is still up in the air. The use of feature extraction techniques determines the degree of recognition accuracy that may be attained and the decrease of misclassification of the two iris layouts. The effective iris recognition framework anticipates accurately capturing the isolated data present in the iris sample. In this document, feature extraction approaches are improved and put into practice. These methods make use of wavelet filters. To represent the biometric template, binary code is applied to the data encoded using the Haar wavelet transform. To categorize iris templates and get the false acceptance rate (FAR), false rejection rate (FRR), and recognition rate, use the Euclidean distance (RR). An effective feature extraction method is the wavelet transform utilizing HAAR. As a result, the two databases used are more likely to be recognized and FAR and FRR are equalized. The FAR and FRR are 1.053 percent and 97.89 percent of anomalies are detected when the HAAR filter is applied to the MMU and UBIRIS v1 datasets.