Study On Information Hiding Algorithm Based On Intra Prediction Modes For HEVC Video

With the rapid development of network and multimedia technology, the application of digital multimedia information becomes more and more popular, and the information security issues become increasingly prominent. Piracy in video, image, and software has a serious impact on the relevant industries and market order. Even worse, illegal use of multimedia works, forgery, or even malicious attacks, makes the integrity and authenticity of information being questioned. In this context, information hiding technology has become an effective method for secret communication and copyright protection. It can effectively curb the illegal use of digital multimedia. Digital video information hiding technology is important and its application is broad. The video files are usually in the form of compressed for storage and transmission because of the great original data. HEVC is the latest video compression coding standard. Researching the video information hiding algorithm has high practical significance and theoretical value under the HEVC standard.This paper illustrated the research background of information hiding technology, significance and research status at home and abroad. Then the video information hiding technology was introduced in detail, including the technical features, classification and evaluation standard. And we do a deep analysis of HEVC codec technology on the basis of the related research, mainly includes the following several parts.Firstly, we analysis the HEVC video codec technology, mainly for HEVC codec framework, the key technologies such as inter prediction coding, quantization and entropy coding. Furthermore, we elaborate the intra prediction coding differences analysis between HEVC standard and H.264/AVC standard, which is based on the information hiding for the next work to lay the foundation.Secondly, we proposed an information hiding algorithm for HEVC based on the parity of the intra prediction modes. We select the 4×4 luminance blocks to embed information with the LBP standard deviation threshold of the pixel values. It reduces the influence of the modulation prediction modes on the quality of the video. The secret information is embedded itself as a key to select location before embedding the information. Depending on the coding patterns of intra frame with Lagrangian model can be decided whether to embed secret information. For reducing the impact of modifying the prediction mode on the quality of video, the best prediction mode will be replaced with the substitute mode which is the one with the least Lagrangian cost among those matching parity with the secret information. The extraction of information merely requires decode the prediction mode from the bit stream and meet the real-time requirements.Thirdly, an information hiding algorithm for HEVC based on intra prediction and block codes is proposed. In the first phase, we establish a(4, 3) standard array decode table with the principle of the standard array. Then we modify the intra prediction modes in 4×4 luminance blocks according to the mapping between the matrix and the prediction mode. Finally, re-encode the changed block with the substitutive prediction mode. 3 bits can be embedded in four 4×4 luminance blocks and 1.25 prediction modes are modified on average, which reduce the impact of the prediction mode modulation on the quality of the video. Experimental results show that the bit-rate increases about 1% and the PSNR loss is just about 0.05 dB. The proposed hiding algorithm can effectively embed information with little influence on the bit-rate and quality of the cover video.Fourthly, we proposed two information hiding algorithms for HEVC based on prediction modes differences. In the first method, intra prediction modes are mapped to the angle values. We are establishing a mapping relationship table between the angle differences and secret information. Modifying the intra prediction modes to embed secret information is based on the mapping. Two bits of information is embedded in two 4×4 luminance blocks and only modify one prediction mode, which reduce the impact of the prediction mode modulation on the PSNR and bit-rate. In the second method, we modify the difference of prediction modes to embed information with the method of pixel value difference. Three bits are embedded in two prediction modes, which improve the capacity of information hiding. Experimental results show that the two proposed methods bring little influence on distortion to the video.Furthermore, we proposed an information hiding algorithm with high capacity for HEVC. We modify the intra prediction modes in 4×4 luminance blocks to embed secret information. According to the probability distribution of the statistical optimal prediction mode and suboptimal prediction mode, we are establishing a mapping relationship between prediction modes and secret information. Modifying the intra prediction modes to embed secret information is based on the mapping. The extraction of information merely requires decoding the prediction mode from the bit stream. Experimental results show that the proposed hiding algorithm can embed information with high capacity and effectively guarantee the subjective and objective quality of video.Finally, in order to validate the proposed information hiding algorithm for HEVC based on intra prediction modes, a video information hiding system is developed. The secret information can be embedded in ‘yuv’ format video files with HEVC coding standard for compression coding. In addition, transparency and real-time of algorithm can be measured using bit-rate, PSNR, encoding time, etc.