Research Of Encrypted Face Image Robust Recognition Algorithm

Recently.face recognition has been widely used in various fields,ranging from the mobile phones which support face recognition,the security checkpoints for train stations and the real-time alert systems for which enhance security during two sessions...In the age of the Internet,face recognition is everywhere.However,information security and privacy protection have also become an indispensable topic in the Internet era.How to ensure the information security is the focus of current research.At present,kinds of face recognition algorithms are performed in the plaintext domain,the security of face image information is often overlooked,so how to solve the problem of encrypted face recognition in the cloud environment is important.Considering the above factors,this paper has carried out the algorithm research of cipher face recognition in the following aspects:The algorithm of face recognition based on DCT,DFT and DWT-DCT in the cloud environment is studied.The feature vector of the encrypted face image is extracted by DCT,DFT and DWT-DCT transform,then matches the feature vectors of the encrypted face image stored in the cloud and the feature vector is recognized as face image“Keyword",finally according to the value of the NC recognizes the face image.According to the experiments,it is verified that the algorithm proposed in this paper not only has ideal homomorphic encryption features,but also has good robustness against conventional,geometric,illumination and mask attacks,and it can also ensure the privacy and safety of face recognition;Finally based on the data of experimental,the robustness performance comparisons of the above three different transform encrypted face image recognition algorithms are compared.We prove which kind of transform encrypted face image recognition algorithms is more suitable under certain conditions,and propose the prospects of encrypted face image recognition.