摘要： 针对H.264/AVC视频压缩标准，提出基于运动矢量分量差的隐写算法。根据人眼视觉特性，筛选出人眼较不敏感的运动矢量，计算其分量差以控制嵌入操作的位置，并在选定的运动矢量分量中嵌入秘密信息。由于操作具有对称性，因此对视频质量的影响较小。理论分析和实验结果证明，该算法具有较高的载体利用率和嵌入效率，能够达到视频隐蔽通信对隐蔽性和隐写容量的要求。

关键词: H.264/AVC标准, 视频隐写, 运动矢量分量差, 盲提取, 视觉不可见性

Abstract: This paper proposes a new video steganography algorithm based on the differences of motion vector components for H.264/AVC video compression standard. In view of the characteristics of Human Visual System(HVS), the algorithm filters the motion vectors which HVS is less sensitive to. It uses the difference between the chosen motion vectors components to control the operating position, and embeds secret message in the component. During embedding, only to modify the component with larger absolute value, and the operation is symmetric, so that the impact of the carrier’s statistical properties is reduced. Theoretical analysis and experimental results show that the algorithm obtains the higher carrier utilization and embedding efficiency, and achieves the requirements of covert communication about the invisibility and capacity.

Key words: H.264/AVC standard, video steganography, difference of motion vector components, blind extraction, visual invisibility

中图分类号: 

• TP309