计算机工程与应用 ›› 2013, Vol. 49 ›› Issue (11): 72-75.

• 网络、通信、安全 • 上一篇    下一篇

基于MCM和HVS的彩色图像盲隐写分析算法

冯  帆1,2,王建华1,张政保2,王惠萍1,戚红军1   

  1. 1.白求恩军医学院 基础部,石家庄 050081
    2.军械工程学院 计算机工程系,石家庄 050003
  • 出版日期:2013-06-01 发布日期:2013-06-14

Blind steganalysis algorithm for JPEG color images based on MCM and HVS

FENG Fan1,2, WANG Jianhua1, ZHANG Zhengbao2, WANG Huiping1, QI Hongjun1   

  1. 1.Foundation Department, Bethune Military Medical College, Shijiazhuang 050081, China
    2.Department of Computer Engineering, Ordnance Engineering College, Shijiazhuang 050003, China
  • Online:2013-06-01 Published:2013-06-14

摘要: 针对典型算法在描述DCT域相关性方面存在的不足,所引起的隐写检测综合性能不理想的问题,将MCM与HVS进行有机结合,提出一种JPEG彩色图像通用隐写分析算法。设计8-邻域MCM型,全面描述DCT系数之间的相关性,基于HVS分别在YCbCr模型空间Y分量抽取Markov状态转移矩、在其相应RGB三通道抽取Markov状态转移矩的主对角线邻域相似熵作为特征统计量,并进行合理“绑定”,采用PCA技术对其进行优化选择,构建高效分类特征向量。实验结果表明该算法对于Jsteg、F5、Outguess、MB1、MB2攻击具有较高的可靠性和检测正确率。

关键词: 马尔科夫链模型(MCM), 人类视觉系统(HVS), 彩色图像, 隐写分析

Abstract: Aiming at the problem that comprehensive performance of typical steganalysis isn’t perfect, which is caused by a lack of characterizing relativity in DCT domain, a blind steganalysis algorithm for JPEG color images is proposed fusing MCM and HVS. It devises an 8-neighborhood MCM model to describe comprehensively the distribution characteristics of DCT coefficients. The model is used for extracting sensibility features statistic based on HVS, which are Markov transition probability matrices of the Y component in the model space of YCbCr, and the main diagonal similitude entropy of the corresponding RGB of Y component, then these feature statistics are in reason merged. PCA technology is applied to optimize original features, and construct effective classified features eigenvector. The experimental results show that the presented method can achieve higher reliability and accuracy rate, aiming at Jsteg、F5、Outguess、MB1、MB2 steganography.

Key words: Markov Chain Model(MCM), Human Visual System(HVS), color images, steganalysis