Study On H.264/AVC Video Information Hiding Technology

With the advent of the big data era, the influence of networks on people’s work and life is growing. The improvement of the Internet and wireless networks offers consumers worldwide ubiquitous channels to deliver and to exchange multimedia information. So the multimedia security should not be ignored. The multimedia data-hiding technique has become one of the most useful methods to protect multimedia by embedding the important information into the redundant signals of the digital media.The digital video has much spatial and temporal redundancy. So it can be considered as a good candidate cover for hiding data. With the development and application of the video compression technology, the involved content of video steganography becomes wider and wider. Nowadays designing steganography algorithms with high concealment, strong robustness and high security is not the only research emphasis. More practical and more distinctive algorithms can be designed.The compressed-domain steganography algorithms based on H.264/AVC have been studied in this thesis. The main work is as follows:A lossless information hiding algorithm based on H.264 inter-prediction is proposed. Secret data are embedded into quantized DCT coefficients of luma residuals in inter-macroblocks by using the histogram lossless data hiding technique. The usefulness of the double-layer embedding method makes the hiding payload higher. The experimental results show that this algorithm can obtain good invisibility and high hiding capacity. In addition, the secret information can be extracted correctly and the original video signals can be recovered.A drift compensation data hiding algorithm based on H.264/AVC is improved. A quinary digit is embedded into two quantized DCT coefficients of luma residuals in intra 4×4 macrolocks by using the diamond coding technique and the coupled relationship of integer transform. The experimental results show that the improved algorithm can hide data without intra-frame distortion drift.And this algorithm can also improve the embedding efficiency and offer a larger embedding capacity in the case of high stego-video quality.