Comparing the Legendre Wavelet filter and the Gabor Wavelet filter For Feature Extraction based on Iris Recognition System

Iris recognition system is today among the most reliable form of biometric recognition. Some of the reasons why the iris recognition system is reliable include; Iris never changes due to ageing and individual can be recognized with their irises from long distances up to 50m away. The iris recognition system process includes four main steps. The four main steps are; iris image acquisition, preprocessing, feature extraction and matching, which makes the processes in recognizing an individual with his or her iris. However, most researchers recognized feature extraction as a critical stage in the recognition process. The stage is tasked with extracting unique feature of the individual to be recognized. Different algorithm over two-decade has been proposed to extract features from the iris. This research considered the Gabor filter, which is one of the most used and Legendre wavelet filters. We also apply them on three different datasets; CASIA, UBIRIS and MMU databases. Then we evaluate and compare based on the False Acceptance Rate (FAR), False Rejection Rate (FRR), Genuine Acceptance Rate (GAR) and their accuracy. The result shows a significate increase in recognition accuracy of the Legendre wavelet filter against the Gabor filter with up to 5.4% difference when applied with the UBIRIS database.