Research Of Iris Recognition Based On Principal Phase Congruency

In the era of rapid development of computer and network technology, biometric technology is increasingly being valued by researchers from various walks of life. Compared with other biometric identification technologies, the texture of iris is rich and stable; meanwhile, Iris recognition technology has greater stability and accuracy, non-invasive, as well. As an emerging biometric identification technology, iris recognition draws more and more attention of the public, it will have broad prospect of application and it is of great study value. This thesis focuses on iris identification algorithms and has made some achievements-The main work is summarized as follows:First, this thesis focuses on the algorithms of iris image pre-processing, on this basis, some improved algorithms were proposed. Histogram statistics and hough transform were used t to achieve fast positioning of iris, this algorithm reduced the amount of computation. To avoid the defects of hough transform and generalized least squares, a combined-line approach was adopt for fitting eyelids.The preprocessed images are substantially free of the eyelids and eyelashes and other noise.Secondly, this article introduces several common iris feature extraction algorithms. The principles phase congruency and log Gabor filter are focused on.Finally, an algorithm of Iris recognition based on principal phase congruency is proposed based on the existing algorithms, which is the innovation of this thesis. In the last, design and implement a system of iris recognition based on principal phase congruency. By comparing with other algorithms, the result shows that the algorithm is more effective.