Improved Algorithm Of Iris Recognition Based On Central Symmetric LBP Operator

In recent years,biometric technology has become more and more popular.Biometric technology is a kind of technology which is based on the biological characteristics of the human body,including iris,fingerprint,palmprint,DNA and walking gait.Compared with other biological characteristics,iris has been paid more and more attention by academic and business circles because of its uniqueness,stability,security and non invasion.Iris payment based on iris recognition technology can become a new direction of future payment.In this paper,the image processing algorithm of iris recognition in iris payment system is described in detail.The main work done is the following:(1)Outlines the basic principle of biological recognition,the main aspects of biological recognition contains the advantages and disadvantages of several indexes;physiological structure of the iris and iris recognition has made the corresponding description;a detailed statement of the test used in the database: CASIA-Iris V4.The CASIA-Iris V4 database consists of six different sub databases,each of which has a different collection,acquisition,and acquisition environment.This paper briefly introduces the object and condition of a complete iris payment system.(2)In the preprocessing of iris image,an edge detection algorithm based on boundary gradient enhancement is proposed to overcome the shortcoming of the outer boundary of iris.Firstly,the pupil gray characteristic is relatively low,the use of binarization to extract the inner edge of the iris image,then use the method of boundary gradient enhancement algorithm and Canny operator combined to extract the iris outer boundary,using the least squares fitting of circle fitting method for internal and external boundary.(3)In the iris feature extraction and matching,we first describe the traditional LBP operator,and introduce the LBP rotation invariant model and LBP equivalent model derived from the traditional LBP model in the paper.In section(2)points on the basis of,this paper used the improved CS-LBP operator was used to extract the feature of CS-LBP operator encoding is discussed,the improved CS-LBP operator can greatly reduce the memory consumption,significantly improve the rate of feature extraction.Finally,using the Hamming distance to match two images.