云环境下基于秘密共享的图像安全检索方案

Privacy-preserving image retrieval scheme based on secret sharing in cloud environment

针对传统云环境下的加密域图像检索方案仅考虑使用单个服务器提供检索服务的问题,提出一种基于秘密共享的图像安全检索方案,使用秘密共享技术构建了一个图像安全检索模型.在预处理阶段,图像所有者生成并加密图像索引的多个秘密份额,并将加密后的索引份额与加密图像一起外包给不同的云服务器;在查询阶段,用户生成并加密多个查询陷门,然后将其分别外包到不同的云服务器,云服务器结合安全多方计算技术,实现图像安全检索.实验结果表明该方案能实现较高的检索精度和安全性.

Aiming at the problem that the encryption domain image retrieval scheme under the traditional cloud environment only considers the use of a single server to provide retrieval services, a secure image retrieval scheme based on secret sharing was proposed, and an image secure retrieval model was constructed using secret sharing technology. In the preprocessing stage, the image owner generated and encrypted multiple secret shares of the image index,and outsourced the encrypted index shares together with the encrypted image to different cloud servers. In the query phase,the user generated and encrypted multiple query trapdoors,and then outsourced them to different cloud servers.The cloud server combined secure multi-party computing technology to achieve secure image retrieval.Experimental results show that this scheme can achieve higher retrieval accuracy and security.