Research On Privacy-oriented Iris Image Processing Method

With the development of the Internet,people pay more and more attention to information security.Biometrics recognition technology protects personal information by analyzing human's own physiological or action characteristics.In recent years,in the field of public safety,biometrics such as fingerprints and human faces have been widely used.However,there are still some limitations in the authentication process.The uniqueness and stability of iris recognition make up for these shortcomings.A series of international studies have shown that iris recognition is currently the safest and most accurate recognition method.Academia and business circles are paying more and more attention to the identification technology based on iris feature extraction.Iris recognition is increasingly used in certain departments that require high security performance,such as banking systems and security agencies,and in some environments that require high authentication.The improved security of iris recognition plays an important role in maintaining social and financial security.This article elaborates on the research background and current status of iris recognition technology,iris recognition technology and security issues in iris recognition process.This paper uses differential privacy to hide iris features.The Gaussian mechanism and the Laplacian mechanism are applied to the iris image to protect the privacy information of the iris features and achieve good iris feature hiding effect.Specifically,when the privacy requirement parameters are given,Laplacian mechanism and Gaussian mechanism are used to hide the iris biometrics in the image.As noise parameters increase,the level of privacy protection becomes stronger.The two mechanisms are compared in terms of algorithm efficiency,and the image quality is evaluated by an image hash algorithm.Experiments have confirmed that the differential privacy protection method can significantly improve the security of iris images.The effects of these two mechanisms are verified by simulation.Simulation results show that under the same noise parameters,the privacy protection based on Gaussian noise is slightly better than Laplace noise.The iris image has the characteristics of large amount of information and high redundancy,making the traditional text data encryption algorithm not suitable for iris image encryption.Due to the characteristics of high initial sensitivity and strong dispersion,chaotic systems are often used in the field of information security,and have achieved extraordinary results.In order to overcome these shortcomings,this paper constructs a relatively complex and difficult to crack double chaotic image encryption algorithm.Only the iris area of the iris image is encrypted,which reduces the computational complexity of encryption and decryption and further improves the operation efficiency.This thesis contains 38 figures,1 tables and 78 references.