基于滤波攻击的高阶统计量隐写分析新方法  

A NEW STEGANALYSIS OF HIGHER-ORDER STATISTICS BASED ON FILTER-ATTACK

作　　者：魏盛昕[1] 林家骏[1] 戴蒙[1] 

机构地区：[1]华东理工大学信息科学与工程学院,上海200237

出　　处：《计算机应用与软件》2008年第6期257-259,共3页Computer Applications and Software

摘　　要：提出一种在小波域中基于滤波攻击的有效的隐写分析方法,首先对待测图像进行滤波攻击,依次在原图和滤波后图像的小波域中计算各自的特征统计量;然后将两者之差作为本文所用的特征统计量,共构造了45维的特征向量;最后采用判别对待测图像进行训练分类并与Farid的方法进行比较,获得了很好的效果。In this article it proposes an effective steganalysis method based on fiher-attack in wavelet-domain. Firstly a filter-attack was made on target image, each feature statistic in wavelet- domain of the original image and the attacked image was computed in turn; then the differences between them were taken as the feature statistics in this method and totally 45-D feature vectors were constructed; at last Fisher Linear Diseriminant （FLD） was used to train and to classify the target image. Compared with Farid＇s method, the new method performs well.

关 键 词：滤波攻击 高阶统计量 隐写分析 

分 类 号：TP309[自动化与计算机技术—计算机系统结构] TN911.7[自动化与计算机技术—计算机科学与技术]