基于运动矢量相位角和卷积码的大容量视频隐写算法

Big-capacity video steganography based on motion vector phase and convolutional code

为提高视频隐写中秘密信息的嵌入容量,提出一种基于运动矢量相位角和卷积码的大容量视频隐写算法。通过研究视频流中每一个帧组里的P帧和B帧上的运动信息,将秘密信息经过交织变换,用相位角的值表示不同的数字序列并作为卷积码的基本生成矩阵,使用卷积码来嵌入秘密信息。实验表明,该算法具有嵌入容量大、不可见性好和稳健性强等特点,在保持良好视频质量的前提下,可以达到视频隐写中大容量嵌入的目的。

In order to increase the capacity of secret information in video steganography,a big-capacity video steganography algorithm was proposed based on motion vector phase and convolutional code.By studying the moving information of P frame and B frame in every Group of Pictures（GOP）,the interlace switch was used to trade with the secret information at first,and then the different size of motion vector phase was used to represent different information to denote the basic generator matrix and the convolutional code was used to insert secret information.The experimental results show that the proposed algorithm not only has a big capacity of insert information,but also has a good imperceptibility and great robustness for secret information.It can achieve high capacity in video steganogrphy and maintain good video quality as well.