The Research Of Fast Identity Recognition Algorithm Based On Iris Recognition Technology

The human iris is vary from person to person.It has life-long stability,the non-contact acquisition make it easy to compute,and the high security make it difficult to counterfeit.All of these advantages make the biometric technology based on iris recognition technology is considered as the most research significance and market potential of identity technology,it has been widely used in national defense,the financial sector and access control system etc.The research work of this paper focuses on the improvement of the three parts of the iris image localization,feature extraction and matching recognition,and analyse the experimental results.,finally obtains the good result.The main content of this study and the achievements are as follows:(1)According to the characteristics of the traditional segmentation algorithm of iris localization efficiency is not high,this paper proposed iris localization algorithm based on combining the extraction of pupil area of interest and partial adjustment,first extract the pupil part of the region of interest by O.stu operator,after binarization processing and partial adjustment,finally achieve precise positioning the iris inner boundary by using Hough transform.When the outer boundary is located,the outer boundary of the iris is searched in the relatively small range by using the coupling relation between the inner and outer boundaries of the iris and the Hough transform.Experimental results show that the proposed algorithm improves the positioning speed and accuracy,and achieves good results.(2)In the feature extraction stage,the CS-LBP operator does not consider the influence of the central pixel on the extraction of iris texture information,this paper proposes an improved algorithm for the CS-LBP operator,that is,the CS-LDP operator.The operator considers whether the region is valid texture information in view of whether the gray value of the two sampling points in the neighborhood symmetry of the center of the image center is symmetrical and whether the gray value of the central pixel has the directionality.It solves the problem that the central pixel is ignored,which makes the iris feature information more comprehensive.The experimental results of the CASIA-V3.0 Interval iris database provided by the Chinese Academy of sciences are presented.The results show that the proposed algorithm can get better results.(3)In recognition,this paper combines the advantages of Support Vector Mach ine(SVM)and Hamming Distance(HD),and proposes a pattern matching algorith m based on SVM and HD.The experimental results show that compared with the t raditional matching recognition algorithm,the correct recognition rate of this algorit hm is 99.955%,which is in accordance with the requirement of most occasions in practical application.