针对LSB的RS隐写分析研究及改进

Research and improvement on RS steganalysis against LSB

介绍了基于数字图像的信息隐藏技术的基本内容及方法,着重研究了图像信息隐藏的反向检测技术——隐写分析技术,总结了其方法和分类。针对隐写分析中的RS(Regularand Singular groups method)检测算法进行了深入的研究,提出了改进算法,在一定程度上减小了RS检测算法对统计假设的依赖性,并且在不改变像素间相关特性的同时可以对图像进行二次检验,增加了检测结果的准确性。在不同的掩码和嵌入率下进行了大量实验,实验结果验证了该改进算法的有效性,基于该改进算法编写了检测程序,实现了对灰度及彩色图像的LSB(Least Significant Bit)隐藏检测。

This paper introduces the steganography and its basic approach on digital image, does further research on its reverse detection technology—steganalysis, including its methods and classifications. It Presents an improvement on RS detection method which to some extent reduces the dependence on statistical hypothesis and increases the detection accuracy of RS by retesting the images without changing the correlation between pixels. The improved algorithm is verified effective under massive experiment data based on different mask and embedding rate. Then it successes in steganalysis on LSB steganography in gray and color images by writting detection program.