计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (4): 93-96.DOI: 10.3778/j.issn.1002-8331.2011.04.026

• 网络、通信、安全 • 上一篇    下一篇

基于PCA和希伯特包络分析的盲隐写分析算法

冯 帆1,2,王嘉祯1,刘会英1,王惠萍2,郭景涛2,张 斌1   

  1. 1.军械工程学院 计算机工程系,石家庄 050003
    2.白求恩军医学院 计算机教研室,石家庄 050081
  • 收稿日期:2009-05-26 修回日期:2009-07-15 出版日期:2011-02-01 发布日期:2011-02-01
  • 通讯作者: 冯 帆

Blind steganalysis algorithm based on PCA and Hilbert envelope analysis

FENG Fan1,2,WANG Jiazhen1,LIU Huiying1,WANG Huiping2,GUO Jingtao2,ZHANG Bin1   

  1. 1.Department of Computer Engineering,Ordnance Engineering College,Shijiazhuang 050003,China
    2.Computer Staff Room,Bethune Military Medical College,Shijiazhuang 050081,China
  • Received:2009-05-26 Revised:2009-07-15 Online:2011-02-01 Published:2011-02-01
  • Contact: FENG Fan

摘要: 为克服传统算法采用离散小波变换(Discrete Wavelet Transform,DWT)最终引起的加性噪声隐写图像检测正确率较低而虚惊率较高的问题,提出了一种新的盲隐写分析算法。综合应用主成分分析(Principal Component Analysis,PCA)和希伯特包络分析的优点,提取隐写图像高频子带希伯特包络解析信息号的主成分特征统计量,计算主成分信息熵,构建信息熵敏感特征向量。采用非线性支持向量机(Support Vector Machine,SVM)分类器,基于Matlab7.1平台进行仿真研究,结果表明:该算法对于空域和DWT域隐写检测具有较高的检测正确率和较低的虚惊率。

关键词: 主成分分析(PCA), 包络分析, 信息熵, 隐写分析

Abstract: A new blind steganalysis algorithm is proposed based on PCA and Hilbert envelope analysis to reduce false rate and improve accuracy of steganography images with additive noise.It extracts principal component feature statistic of stego-images envelope signals,calculates information entropy of principal component,and constructs sensitivity feature vectors of information entropy using the advantage of PCA and Hilbert envelope analysis.The simulation results based on Matlab7.1 using nonlinear SVM show the presented method can achieve higher accuracy rate and lower false rate,aiming at steganographic on spatial domain and DWT domain.

Key words: Principal Component Analysis(PCA), envelope analysis, information entropy, steganalysis

中图分类号: