Efficient Privacy-Preserving Content-Based Image Retrieval in the Cloud

Cloud storage systems are increasingly being used to host personal or organizational data of critical importance, especially for image data that needs more storage space than ordinary data. While bringing in much convenience, existing cloud storage solutions could seriously breach the privacy of users. Encryption before outsourcing images to the cloud helps to protect the privacy of the data, but it also brings challenges to perform image retrieval over encrypted data. To address this issue, considerable amount of searchable encryption schemes have been proposed in the literature. However, most existing schemes are either less secure or too computation and communication intensive to be practical. In this paper, we propose an efficient privacy-preserving content-based image retrieval scheme. We first convert the high-dimensional image descriptors to compact binary codes, and then adapt the asymmetric scalar-product-preserving encryption (ASPE) to ensure the confidentiality of the sensitive images. The security analysis and experiments show the security, accuracy and efficiency of our proposed scheme.