Research On Iris Recognition Algorithm Based On 2D Log-Gabor Wavelet

In recent years, with the development of information technology, information security has become one of the important research topic in today’s society. Iris recognition technology has many advantages, such as the uniqueness, stability, high recognition rate, non invasive. Because of these, it became the hot research filed at present. In the study of the iris recognition, researchers have proposed some algorithm. This paper puts forward some ideas on the edge of the iris localization and iris feature extraction. The main work is as follows:1. In the aspect of image preprocessing:Using the improved Sobel operator to enhance the changed point and line in the picture. With the method of circular symmetrical and Daugman, I can locate the inner and outer edge of the iris positioning and accurate positioning. Using a Log-Gabor filter to test the eyslash.2. In the aspect of feature extraction:Gabor filter is not sensitive to light reflection, so it can’t remove light on texture image perfectly. The quality of iris image is affected, thereby affect the recognition rate. Aiming at the shortcomings of the Gabor filter, this paper uses the Log Gabor filter in extracting feature and analysis the influence of the selection of filter parameter on the feature extraction accuracy. This paper makes up for the deficiency of Gabor filter and get very good recognition effect. Using the hamming distance to match iris texture feature vector and put the application implement of feature extraction finally. Then the results show that the algorithm improves the recognition speed in the case of a higher recognition accuracy.Do experiment on the data CASIA, the selection of hamming distance is 0.355. The recognition tate on CASIA-IrisV3-Interval is 98.7%, and the error rate is 0.038%. On the Lamp database, the recognition is 98.3, and error rate is 0.18%. The experimental results show the proposed algorithm has better versatility and stability.