一种基于相邻像素差值的隐写分析方法

Steganalysis Method Based on the Difference of Adjacent Pixels

首先介绍BMP文件结构及特点,基于相邻像素相关的特性,提出了一种新的针对LSB替换的隐写分析算法.分别从横向和纵向扫描图像像素并统计相邻像素之间差值,然后根据差值奇偶和的比值对图像是否含有隐秘信息进行判断.实验编程环境为VS2008,实验结果表明此检测算法统计量小、实现简单,在隐秘信息嵌入量较多时具有较高检测率.

This dissertation first introduces the structure and characteristic of BMP file. Based on the features of adjacent pixels, a steganalysis algorithm for the replacement of LSB is proposed. Scanning and counting the difference of adjacent pixels horizontally and vertically are the first step, and then this thesis judges whether the image contains secret information according to the ratio of odd and even. The development environment is VS2008, and the experiment shows that the algorithm can be simply applied with a small statistics. In addition, the new algorithm has a high detection capability with a large amount of secret information.