Survey on the privacy-preserving content based image retrieval

Abstract

With the widespread popularity of smart devices and social media, the number of image data is exponentially increasing. The data owners tend to outsource the local data to the cloud servers, where data is stored, shared and retrieved. However, the content of users’ image data contains a lot of sensitive information, which may be exposed to the attackers and the incomplete trusted cloud servers, resulting in a serious risk of privacy leakage to users. The research progress of the content-based image retrieval technology under the privacy-preserving in recent years were reviewed, and the key image cryptographic technology were summarized, including homomorphic encryption, randomized encryption and comparative encryption. Around these techniques, the typical solutions are analyzed and compared in detail, and the improvement strategies of index construction are introduced. Finally, the future research directions are discussed.

Keywords

• privacy-preserving
• content-based image retrieval
• symmetric searchable encryption
• cloud computing security