加密
计算机科学
图像共享
同态秘密共享
秘密分享
明文
计算机安全
图像(数学)
无损压缩
理论计算机科学
密码学
计算机视觉
数据压缩
作者
Rui Wang,Guozheng Yang,Xuehu Yan,Shengyang Luo,Qi Han
标识
DOI:10.1016/j.jvcir.2023.104013
摘要
Nowadays, digital images have a greatly widespread and profound influence on human life, and its security protection problem is increasingly prominent. Image encryption is one of the most effective solutions for image protection but may result in secret concentration or single point of failure. Secret image sharing (SIS) with loss tolerance provides a feasible method to protect secret images in a distributed way. However, existing SIS schemes are most utilized in natural images, and the application of SIS to encrypted images which are widely used in practical scenarios is rarely discussed. Thus, this paper proposes a novel concept of SIS in the encrypted domain (SIS-ED) for the first time on the basis of synthesizing the merits of SIS and image encryption. SIS-ED integrating the encryption and SIS means applying general or user-defined encryption algorithms to encrypt the secret image in the plaintext domain, and in the encrypted domain using appropriate SIS schemes on the encrypted image to share secrets. SIS-ED can prevent secret images from being reconstructed illegally and enhance security against malicious attacks such as differential attacks, which can be widely applied in practical scenarios with high-security requirements such as authentication, blockchain, and cloud systems. In addition, we present a concrete SIS-ED scheme to instantiate the SIS-ED model. The proposed SIS-ED scheme with high security can realize lossless recovery without pixel expansion, and reduced-size shares. Theoretical analysis, experiments, and comparisons are conducted to show the effectiveness and advantages of SIS-ED.
科研通智能强力驱动
Strongly Powered by AbleSci AI