一种图像隐写分析方法

doi:10.3969/j.issn.1006-2475.2011.11.019

计算机与现代化 ›› 2011, Vol. 1 ›› Issue (11): 70-4.doi: 10.3969/j.issn.1006-2475.2011.11.019

一种图像隐写分析方法

张文华¹，向阳¹，潘天奎²

1.西安通信学院，陕西西安710106； 2.解放军73683部队，福建三明353304

收稿日期:2011-07-15 修回日期:1900-01-01 出版日期:2011-11-28 发布日期:2011-11-28

A Method of Images Steganalysis

ZHANG Wen-hua¹, XIANG Yang¹, PAN Tian-kui²

1.Xi’an Communications Institute, Xi’an 710106, China； 2.PLA 73683 Troops, Sanming 353304, China

Received:2011-07-15 Revised:1900-01-01 Online:2011-11-28 Published:2011-11-28

摘要/Abstract

摘要：

针对现有隐写分析算法对彩色图像的检测效果不理想的问题，提出一种基于小波统计特征的隐写分析算法来分析嵌入信息前后载体图像与隐写图像之间的统计特征的差异。该算法首先将图像变换到YCbCr空间，其次通过小波分解提取预测子带的小波系数概率密度函数矩(PDF)和预测误差图像的子带系数直方图特征函数矩(CF)，最后运用支持向量机(SVM)进行分类。实验表明，该方法对于流行的F5、OutGuess和Jphide&Seek隐写方法具有较强的检测能力。

关键词: 隐写分析, 预测误差图像, PDF矩, CF矩

Abstract:

As for the deficiency of the existed steganalysis algorithm for color image detection, this article provides a steganalysis algorithm based on wavelet statistics to analyze the statistic differernce between carrier image and stego image. The method, transforming the color image into the YCbCr space and extracting the prediction subband coefficients’ probability density function(PDF) and the predictionerror image subband coefficients’ histogram characteristic function(CF) moments through the wavelet decomposition, makes use of the support vector machine(SVM) to achieve classification. The experimental results for the popular JPEG steganographic schemes, for example, F5, OutGuess, Jphide&Seek, demonstrates that the proposed scheme can achieve good detection accuracy.

Key words: steganalysis, predictionerror image, PDF moments, CF moments

张文华;向阳;潘天奎. 一种图像隐写分析方法[J]. 计算机与现代化, 2011, 1(11): 70-4.

ZHANG Wen-hua;XIANG Yang;PAN Tian-kui. A Method of Images Steganalysis[J]. Computer and Modernization, 2011, 1(11): 70-4.

一种图像隐写分析方法

A Method of Images Steganalysis

可视化

被引次数

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 0

编辑推荐

Metrics

本文评价