Research On Steganography Algorithm For3D H.264Video Without Distortion Drift

As one of the important ways to realize covert communication, video steganography technology has become an important research area in the field of security and academia.3D H.264video will be the main mass media in the future. The secret communication based on3D H.264video will be a special battlefield for the world’s information warfare and network warfare. The rapid development of the3D H.264video brings new opportunities and challenges to the steganography theory and technology. It is a new research field of studying how to realize the video steganography algorithm based on3D H.264video with high imperceptibility, strong robustness and reversibility. Research on these new problems, difficulties and key problems, even in the field of2D video, has only just begun. These issues not only are the hot research topics today, but also are the cores and foundations of the video steganography based on3D H.264in the future.In order to realize high imperceptibility of the video steganography, with the analysis of3D H.264video coding structure, we have studied and put forward a3D H.264video steganography algorithm without distortion drift based on the inter-frame macroblock. With the analysis of the coding structure of3D H.264videos, we propose a new data hiding algorithm without inter-frame and intra-frame distortion drift. Firstly, an ideal that modify the4×4QDCT coefficients of the inter-MBs except the rightmost column and the bottom row is proposed, and it can completely prevent intra-frame distortion drift. Then, we point out that hiding data into b4-frames can completely prevent inter-frame distortion drift. At the same time, the proposed algorithm does not require to obtain the data of each MB’s MB-type and prediction mode in advance, which greatly reduces the complexity of the algorithm, as well as improves the embedding efficiency. The proposed algorithm has lower complexity and a greater embedding capacity than the current algorithms. Based on nine test video sequences, experimental results confirm the embedding capacity and complexity superiority while keeping the similar human visual effect when compared with current algorithms.Further, with the analysis of the DCT transformation and quantization of the3D H.264video encoding process, a video steganography algorithm without distortion drift based on coefficient compensation for3D H.264is proposed. Instead of perturbing QDCT coefficient-pairs to embed data only into intra-frame4×4luma blocks, this paper allows all the MBs, whose transformation matrix is4×4, including inter-frame and intra-frame, to embed data.The proposed algorithm has solved the problems of pre-reading the original video and inter-frame distortion drift in the current algorithms. The algorithm has lower complexity and a greater embedding capacity than the current algorithms. And the proposed algorithm maintains great randomness by using random sequences. Experimental results confirm the embedding capacity and complexity superiority of the proposed algorithm while keeping the similar human visual effect.On this basis, in order to resist the noise pollution in the process of transmission, we proposed a robust video steganography algorithm for3D H.264based on convolution code technologies and interleaving techniques. In modern communication, Error-Correcting Codes are often used to improve the system performance, among which the convolutional code is adopted widely because of its excellent capability. Combined with interleaving techniques, the proposed algorithm can effectively correct the random errors and burst resulting from passive attacks in the transmission process. At the same time, combined with the distortion drift control techniques above, a robust video steganography algorithm for3D H.264based on convolution code technologies without distortion drift is proposed. And the algorithm can provide theoretical and technical support for covert secure communications under complex network environments. Experimental results confirm the embedding capacity and redundancy superiority of the proposed algorithm while keeping the similar human visual effect.Considering the reversible requirement of video steganography algorithm in some particular field, such as medicine, remote sensing images and military, we proposed a reversible3D H.264video steganography algorithm based on motion vector according to the3D H.264video coding structure. By establishing the inner product between the motion vector and the modulation vector and setting the embedding conditions, we embed1bit data into each motion vector and the proposed algorithm is reversible. Moreover, in order to avoid distortion drift, we only embed data into b4-frames with the coding feature of3D H.264video. Meanwhile, the proposed reversible method has generality and enriches the theory and method of reversible video steganography. Based on twelve test video sequences, experimental results confirm the embedding capacity and complexity superiority while keeping the similar human visual effect when compared with current algorithms.