一种基于统计的LSB隐写分析方法

A Steganalysis of LSB based on Statistics

本文提出了一种基于统计的LSB隐写分析方法,这种方法根据隐写所引起的像素灰度直方图的异常来判断秘密信息的有无,估计了各种情况下的漏检概率和虚警概率。实验结果表明,当秘密信息量较大时,两类错误概率都可以得到有效控制。

This paper proposes a novel steganalytic technique based on the statistics method. This method enable us to judge the presence of a secret message by the change of the histogram caused by information hiding. We estimate the probability of both false and missing a secret message. Numerical results show both kinds of judgment error probability can be effectively controlled with the increase of embedded information.