Iris recognition based on Kaiser filter and whole phase analysis

Kaiser function is used to establish clear edge Kaiser filter channel which has selectivity in the frequency and direction with the characteristic of high adaptability and adjustable performance to feature the iris frequency, which is to modify the filter channel edge faintness due to the spectrum leakage when Gabor filter does the data truncation. Then complete classification of the weighted phase correlation algorithm is used to classify the characteristic of iris phase extracted from Kaiser filter which can reflect subtle difference among the iris characteristics better. Experiment shows that: The accuracy of iris recognition is 99.58% by using the algorithm in this article, while the FAR is 0.12% and the FFR is 0.16%. Thus, the algorithm is good at feature extraction, classification and recognition.