Research On Video Information Hiding Method Based On HEVC

With the development of digital multimedia technology,more multimedia resources are widely used.However,in the process of video transmission and storage,some important information may be illegally edited,embezzled and attacked.Information hiding is a technology that embeds data into media content,mainly including steganography and digital watermarking.This technology has been used in many applications,such as ownership statement,privacy protection and covert communication.In order to meet the requirements of video compression performance brought by the diversification of video applications and the trend of high-definition,High Efficient Video Coding(HEVC)was jointly formulated by the International Telecommunication Union and the dynamic image expert group.On the basis of H.264/AVC,it can improve the compression performance and develop new coding technology,which can be better applied to the development of high-definition and ultra-high-definition video.Therefore,video steganography and watermarking technology based on HEVC has become a research hotspot.By studying HEVC encoding and decoding technology,this paper carries out information hiding based on HEVC coding standard.(1)An information-hiding algorithm based on intra-prediction mode and matrix coding is proposed.Firstly,eight prediction modes are selected from 4×4 luminance blocks in I frame to embed the hidden information.Then,the least significant bit(LSB)algorithm is used to modulate the LSB of the last prediction mode.Finally,the modulated luminance block is re-encoded to embed four bits secret information.Experimental results show that the algorithm improves the embedding capacity,guarantees the subjective and objective quality of the video,and the bit rate increases by 1.14% on average.(2)Combined with the coding method of transformation and quantization process in HEVC,a distortion drift free HEVC adaptive hiding method is proposed.Firstly,the appropriate carrier is selected according to the intra prediction direction and multi coefficient method with distortion drift free to avoid intra distortion drift.Then,the distortion cost function is designed by using the syndrome trellis codes(STC)coding technology.Finally,according to minimizing the distortion cost,the message is embedded into the QDST coefficient of 4×4 brightness block of the selected frame that meets the specific conditions.Experimental results show that the proposed video hiding algorithm can effectively avoid intra frame distortion drift and obtain good visual quality.(3)We propose a semi-blind robust HEVC watermarking algorithm after analyzing the characteristics of HEVC recompression.We discuss the influencing factors about choosing embedding cover,and deduced the texture complexity calculation formulation.Then,the compressed videos are entropy decoded partially to gain the all none-zero coefficients of intra luma blocks.Finally,the binary image information is completely embedded after modulating the signs of the 2 QDST coefficients of the cover.At the same time,the indexes information of watermarking embedding are recorded and transmitted to the decoder as auxiliary information.After analyzing the video subjective quality and video objective quality,there is no obvious artifact after embedding watermark,which reflects the proposed algorithm has good imperceptibility.The results show that the algorithm is highly robust to recompression with the some QP.The embedding process of the proposed algorithm is simple and fast,while the watermark embedding does not make the video bit rate increase,and effectively limiting the degradation of the video perception quality.Figure 32 Table 14 Reference 76...