基于图像子空间和核稀疏表示的人脸识别及同态加密方案设计与研究

Design and Research of Face Recognition and Homomorphic Encryption Scheme Based on Image Subspace and Nuclear Sparse Representation

摘要: 随着大数据和云计算时代的到来，越来越多的用户信息被完全暴露在各大互联网媒体上，这样就存在很大的隐患和安全性问题，为了保护用户的个人隐私信息，完成人脸图像数据在服务器端的隐秘计算，本文提出了核鉴别稀疏保持嵌入算法(KDSPE)结合密码学领域的同态加密和基于身份加密系统(IBE)的不经意传输协议的一种人脸图像隐秘识别方案。终端采集待测样本的数据和数据库的人脸图像数据进行对比，从而判断终端采集的人脸数据是否在数据库存在。这里利用KDSPE算法得到核鉴别稀疏矩阵，然后利用同态加密和基于身份加密系统(IBE)的不经意传输协议来隐秘计算终端和服务器人脸的核鉴别稀疏矩阵的欧式距离，进而判断是否匹配。该算法的优点在于除了可以有效的提取人脸非线性特征外，同时在非约束环境下(姿态，表情，光照，遮挡，年龄，拍摄角度)也有较好的鲁棒性，此外由于结合密码学的知识，该算法还可以保证通信参与者的数据安全和通信通道的安全性，实验结果表明本文算法提高了人脸识别率，同时具有一定的算法安全性。

Abstract: With the arrival of the era of big data and cloud computing, more and more user information is completely exposed on major Internet media. This poses great hidden dangers and security problems. In order to achieve protection of the user’s personal privacy information, completing the secret calculation of face image data on the server side, this paper presents a facial image based on Kernel Discriminative Sparse Keeping Embedded Algorithm (KDSPE) combined with homomorphic encryption in cryptography and inadvertent transfer protocol based on Identity Encryption System (IBE) stealth identification algorithm. The terminal collects the data of the sample to be tested and the face image data of the database to compare, so as to judge whether the face data collected by the terminal exists in the database. The kernel sparse matrix obtained by the discriminative sparse hold embedding algorithm (KDSPE) is used here, and then the Euclidean distances of the nuclear sparse matrices of the human face of the terminal and the server are secretly calculated using the homomorphic encryption and the inadvertent transport protocol based on the identity encryption system (IBE) and then determining whether it matches. The advantage of this algorithm is that it can not only effectively extract facial nonlinear features, but also has good robustness in non-constrained environments (attitude, expression, lighting, occlusion, age, and shooting angle); in addition to the combination of knowledge of cryptography, this algorithm can also guarantee the data security of the communication participants and the security of the communication channel. Experimental results show that the proposed algorithm improves the face recognition rate and has certain algorithm security.