基于图像纹理的空域富模型隐写分析研究

摘要

图像纹理复杂度是隐写分析在攻击自适应隐写时常常关注的问题，而空域富模型(Spatial Rich Model)常用于针对传统非自适应图像空域隐写的攻击，但空域富模型却没有考虑图像纹理的纹理特征。考虑到空域富模型的准确率及图像纹理复杂度对自适应算法的作用，兼顾效率与准确率，本文结合两种思路，将图像纹理复杂度引入到空域富模型中，提出了一种基于图像纹理的空域富模型隐写分析方法。用MATLAB搭建了测试环境进行测试，通过实验结果可以看出，该方法能提高对于自适应隐写算法的检测准确率。 The complexity of image texture is often concerned when steganalysis is used for attacking adaptive steganography, while spatial rich model is often used to attack the traditional image steganalysis. But the spatial rich model does not consider the texture features of image texture. Considering the accuracy of the spatial rich model and the effect of the image texture complexity on the adaptive algorithm, considering both efficiency and accuracy, this paper combines two ideas, introduces the image texture complexity into the spatial rich model, and proposes a steganalysis method based on the image texture. The test environment is built by MATLAB. The experimental results show that this method can improve the detection accuracy of the adaptive steganography algorithm.