The Research On Robustness Methods Of Steganography For H.264/AVC

Data hiding is a kind of technology,which embed the secret information in the digital text, images, audio, video and so on, Which can make the existence of secret information unnoticeable, can not only protect the information content but also the secret behaviors. H.264/AVC, as a new video decoding standard, has become the most common and most effective compression method in the current network video because of its high compression rate, wide applicable range, strong error resilience, etc,.Then,data hiding for2d/3d h.264/AVC has begun in domestic and internation frontier research.H.264/AVC video with secret information transmited on the internet, on the one hand, may be packet loss, frame loss, or bit error, make the secret information cannot recover because of the bad physical environment, on the other hand, some network attacks (eg tampered、replay、re-quantitative or re-coding, etc) and make the secret information cannot be recovered. Therefore, to h.264/AVC video, the high security and strong robustness steganographic methods is the key to the video concealed communication problems.A robust data hiding method using paired-coefficient and BCH code to recover error bit is proposed. We use BCH code to encode the secret information before data hiding. Based on the summary of our former data hiding method, we choose the embedded blocks to embed bits into the paired-coefficients. BCH (63,7,15) is the most powerful code in terms of error correction capability by analysis and experiment, which can correct100%error bits. BCH (7,4,1) is the most powerful code among these several codes in terms of error correction capability when correcting a random error. To BCH (n,k,t), the larger the t value is, the stronger the robustness of the hidden data against the re-quantization and re-encoding attacks is, and viceversa. It is proved analytically and shown experimentally that the proposed algorithm can avert intra-frame distortion drift and achieve high error resilience capacity.A robust data hiding method using single coefficient and BCH code to recover error bit is proposed. Generally, recovering carrier is not considered after extracting embedded data in video Steganographic method, so that the original carrier is destroyed. In this case, this method tries to restore the original carrier while extracting the embedded data. Before data hiding, the hiding data is embedded by BCH to improve the ability of correcting the error bits. Then, we choose specific blocks to control intra-frame distortion drift by using prediction mode and use single coefficient to reduce the modification of the original video. When extracting the embedded bits, we correct errors first, and then try to restore the original video. From results of the experiments, this method is able to restore the original video as far as possible and can protect the original carrier when extracting hiding information.A robust data hiding method is proposed. Data hidding will modify the carrier mostly during the data hiding process, which cannot be restored after the data extracted. In those areas which do not tolerate permanent distortion of the carrier, such as medical imaging and legal evidence, it is unacceptable. We explore the method that restores the original video completely when extracting the secret data. Before data hiding, the secret data is first encoded by BCH. Then we use the prediction mode to select the special embedding blocks to control intra distortion drift. When data hiding, we use the single coefficient embedding method and modify all the coefficients which are greater than the settings. While extracting the secret data, we first correct the error bit, and then restore all the modified coefficients. The experimental results and the performance of the existing methods show that the method has strong robustness and a better visual effect and reversibility.In order to solve frame loss caused by variety reasons, the recovery of lost frames based on (t, n)-threshold secret sharing robust method is proposed. During the embedding process, the embedded secret data is distributed as a specific polynomial. Then we obtain a series of sub-secret. We encode the sub-secret by BCH code and then embed the encoded sub-secret into each frame of the video. During the extraction process, after correcting the error bits by BCH code, we restore the complete hidden data by using the sub-secrets of the video frames, which can achieve the purpose of correcting bit error,frame error, frame loss, and the method has the high frame loss and frame error robustness, when the random frame rate is less than or equal to20%, the method can recover secret data completely.The experiment results show that, in addition to robustness, the method can avoid intra-frame distortion drift and has better visual concealment.