Study On Joint Encryption And Data Hiding Scheme For Images Based On Compressive Sensing

Currently, the application of digital images in the network has been very common. During transmission, digital images are easy to be stolen, copied and tampered; therefore, the security of digital images has received more and more attention.As two main branches of information security technology, cryptography and data hiding realize different functions: encryption transforms the data into unintelligible forms, and only the authorized customer can recover the data correctly; while data hiding embeds some information into multimedia content by modifying the carrier content imperceptibly to authenticate carrier’s ownership or identification etc.. In application of multimedia protection, the joint encryption and data hiding scheme is preferred for two different purposes.As alternative for signal acquisition, Compressive Sensing(CS) allows us to unify sampling and compression. The sampled signal has a certain security and robustness. In this thesis, research and analysis on the existing algorithms of joint image encryption and data hiding have been done and two kinds of CS-based joint image encryption and data hiding algorithms are proposed.The first algorithm is a scheme for separable data hiding in encrypted image based on CS. Firstly, perform operations on DCT coefficients as follows: the stream cipher algorithm(RC4) is applied on the DC coefficients(encryption part); one part of AC coefficients(data hiding part) is reserved and the other is set to be zero for data hiding. Then both the data hiding and encryption of the data hiding part are realized in the CS procedure. In the proposed scheme, the way to encrypt image content and embed additional data is separable and CS not only makes the cipher image compressed but also improves the security of the embedded data. Experimental results show that the proposed method has the merits of high-capacity, high accuracy of information extraction, good performance on image recovery and anti-packet loss.The second algorithm is a joint image encryption and data hiding scheme based on CS. Firstly, the transform domain coefficients of the original image are scrambled by Arnold method controlled by secret key k1. Then, using secret key k2, we realize the unified data hiding and further encryption by CS. Experimental results show that our algorithm achieves good performance in view of security, embedded capacity, extraction accuracy and reconstruction. Not only the data hiding but also the encryption procedure is robust. Simulation results show that the algorithm is robust to packet loss and noise attack.