Cloud-Assisted Image Double Protection System With Encryption and Data Hiding Based on Compressive Sensing

Abstract

In this paper, we propose a compressive sensing(CS)-based scheme that combines encryption and data hiding to provide double protection to the image data in the cloud outsourcing. Different domain techniques are integrated for efficiency and security. After the data holder gets the sample of the raw data, he embeds watermark into sample and encrypts it, and then sends the protected sample to cloud for storage and recovery. Cloud cannot get any information about either the original data or watermark in the CS recovery process. Finally, users can extract the watermark and decrypt the data recovered by cloud directly in sparse domain. At the same time, after extracting the watermark, the image data of user will be closer to the original data compared with the data without extraction. Besides, the counter (CTR) mode is introduced to generate the measurement matrix of CS, which can improve security while avoiding the storage of measurement matrixes. The experimental results demonstrate that the scheme can provide both privacy protection and copyright protection with high efficiency.