Research On Some Key Algorithms In Iris Recognition

As one of the most important biometric technologies,Compared with other biometrics,it has the advantages of uniqueness,anti-counterfeiting and stability,However,iris recognition technology still has many deficiencies in practical application: The acquired iris image is small in size and few in effective pixels.Strabismus,occlusion and blur are easy to occur when acquiring iris images.These factors affect the quality of iris images,the speed and accuracy of iris location,and the accuracy of iris feature extraction and matching,thus affecting the speed and credibility of iris recognition.In view of the above problems,this paper starts with the research of iris image quality evaluation,studies iris location algorithm and feature extraction algorithm,in order to improve the efficiency of iris recognition on the basis of previous work.In this paper,firstly,the iris recognition technology at home and abroad is studied in detail,and the current mainstream iris recognition algorithms are compared and analyzed.The related work in iris image quality evaluation,iris location,feature extraction and matching is stated in detail.The improved algorithm is verified and analyzed based on CASIA?V1.0 database of Chinese Academy of Sciences.The main contents of this paper are as follows:1.Iris quality evaluation.In order to solve the influence of low-quality iris images on iris recognition performance,this paper proposes an iris quality evaluation method based on morphological and gray differences,which removes the low-quality iris images during acquisition and improves the accuracy of iris recognition system.2.Iris localization.Aiming at the problem of low localization efficiency caused by insufficient contour information of iris inner and outer edges,In this paper,a localization algorithm combining least square method and block operator is proposed.Firstly,the outer edge of iris is fitted and located by least square method,then the inner edge of iris is coarsely located by Hough transform,and finally the block operator is used for precise localization,which improves the accuracy and speed of iris localization.3.Iris feature extraction.This paper comprehensively analyzes several feature extraction algorithms,and proposes an iris feature extraction algorithm based on improved 2D-Gabor filter and wavelet transform.The algorithm provided in this paper is verified by Matlab on the CASIA?V1.0database.The algorithm in this paper is compared with other algorithms,and the ideal experimental results are obtained.The validity of the proposed algorithm is proved.feasibility.