Iris recognition using rapid Haar wavelet decomposition

In this paper, we propose an iris recognition system using a basic and fast Haar wavelet decomposition method to analyze the pattern of a human iris. This system has two main modules, which are the feature encoding and iris code matching modules. Among all feature extraction methods, Haar wavelet decomposition is chosen for its computational simplicity and speed in filtering the iris pattern. In the feature extraction module, unrolled iris images are filtered using high pass filter and low pass filter for four times to produce the corresponding coefficients. Subsequently in the second module, Hamming distance between iris codes is calculated to measure the difference between the query iris image and the iris images in the database. Iris recognition is then performed by matching the iris pair with the minimum Hamming distance. This system is tested with 450×60 pixels iris images from the CASIA iris database (CASIA-IrisV3-Interval), and a recognition rate of 98.45% is achieved.