一种基于运动矢量空间编码的HEVC信息隐藏方法

A HEVC Information Hiding Approach Based on Motion Vection Space Encoding

HEVC(High Efficiency Video Coding)是ISO/IEC和ITU-T联合制定的最新的视频编码标准,该文提出了一种基于运动矢量空间编码的HEVC信息隐藏方法.文中首先给出了运动矢量空间的构建及编码方法,并定义了运动矢量集合与该空间中点的映射关系;其后,给出了通过修改运动矢量集合映射值进行信息隐藏的方法.文中的方法实现了在N个运动矢量分量中最多改变一个分量即可嵌入一个2 N+1进制数的效果,具有较高的嵌入效率.此外,文中的方法仅选择CTU(Coding Tree Unit)中尺寸最小的N/2个PU(Prediction Unit)的运动矢量作为嵌入载体,使得隐写所引入的视频附加失真进一步减小并提升了隐写的透明性.最后,通过实验证实了该文的方法可行且具有优良的性能,具有良好的应用前景;而且,该文通过构造虚拟空间进行信息隐藏,给出了一种提升视频信息隐藏算法嵌入效率的新途径.

HEVC is the newest video coding standard jointly developed by ISO/IEC and ITU-T. This paper proposes an information hiding approach based on motion vector space encoding for HEVC video.This paper first introduces the motion vector space construction and encoding methods;then it defines the mapping relationship of motion vector set to the points in the motion vector space.And information hiding is realized by modifying the mapping value of the motion vector set.This method has a better embedding efficiency,because a 2N ＋1-ary number can be embedded by modifying mostly one element of a set with N motion vector components.Furthermore,only the motion vector components from the N/2 PUs with smallest sizes in a CTU can be selected as the secret information carrier,which reduces the video quality distortion caused by information hiding and improves the transparency of steganography.At the end of the paper,the experiments verify that this method is practicable and has excellent performance.Moreover,our method hides information by introducing the virtual space,and it gives a new way to how to heighten the embedding efficiency of video information hiding.