基于同态加的压缩感知加密域信息隐藏算法  

Information hiding algorithm in compressive sensing encrypted domain based on homomorphism addition

作　　者：李名[1,2] 信鑫 LI Ming;XIN Xin(College of Computer and Information Engineering,Henan Normal University,Xinxiang 453007;Key Laboratory of Artificial Intelligence and Personalized Learning in Education of Henan Province,Xinxiang 453007;School of Engineering,Guangzhou College of Technology and Business,Foshan 528000,China)

机构地区：[1]河南师范大学计算机与信息工程学院,河南新乡453007 [2]河南省教育人工智能与个性化学习重点实验室,河南新乡453007 [3]广州工商学院工学院,广东佛山528000

出　　处：《计算机工程与科学》2024年第9期1598-1605,共8页Computer Engineering & Science

基　　金：河南省教育厅2023年度河南省高等学校重点科研项目(23A520009);河南省科技厅2021年度河南省重点研发与推广专项(科技攻关)(212102210413)。

摘　　要：信息隐藏可为云和物联网环境中的海量数据提供必要的安全保护,传统的加密技术虽然有效保护了图像的隐私,但是无法同时提供版权、完整性等方面的保护,因此,在加密域进行信息隐藏面临着较大的需求和挑战。提出了一种在压缩感知同态加密域进行信息隐藏的算法。首先,对压缩感知的同态性进行探索,发现对压缩感知获得的测量值进行加倍,与直接扩展原始信号后再进行压缩感知具有相同的效果。然后,利用同态加运算实现基于差分扩展的压缩感知加密域的信息隐藏。实验仿真结果表明,该算法具有较好的隐私保护性能和信息隐藏性能,并且与最新的加密域信息隐藏算法相比,具有更高的嵌入容量。Information hiding ensures necessary security of massive data in the cloud and Internet of things environments.Although traditional encryption can protect the privacy of the image effectively,it cannot protect the image in other aspects such as copyright and integrity.Therefore,information hiding in encrypted domain is required and challenged.An information hiding scheme in homomorphic encrypted domain is proposed.Firstly,the homomorphism of compressive sensing is explored,and it is found that doubling the measurements obtained by compressive sensing is equivalent to directly extending the original signal before compressive sensing.Secondly,by using homomorphic addition,information hiding in the compressive sensing encrypted domain is realized based on differential extension.The experimental simulation results show that the proposed algorithm has satisfactory performances in both privacy protection and information hiding,and the embedding capacity is higher than the related state of art works.

关 键 词：压缩感知 同态加密 差分扩展 信息隐藏 

分 类 号：TP393.084[自动化与计算机技术—计算机应用技术]