基于图像的信息隐秘检测系统的设计

Design of Blind Detecting System for Image Steganalysis

设计并初步实现了一个基于图像的隐秘检测系统,其目的是在没有原图像载体的条件下,提高发现网络中的隐秘图像的准确率。系统采用对彩色图像每个颜色通道分别进行小波分解,根据小波分解系数绝对值和绝对值线性预测的对数误差生成特征向量,并采用非线性的支持向量机进行模式分类。讨论了该系统的结构、工作原理、控制流程及设计中的关键技术,并对系统性能进行了测试评估,指出了可进一步改进完善的方向。

This paper designs and implements a blind detecting system for images steganalysis. The aim of devising the blind detecting systems is to discover images in networks with more veracity at condition of no original images. The wavelet decomposition is implemented in each color channel, the magnitude of decomposition coefficients and the log error between the actual coefficient and the predicted coefficient magnitudes are used to yield statistics. The non-linear support vector machine algorithm has been employed in the pattern discrimination. Its structure, principle, controlling process as well as the key designing technique of software are also presented in detail. It points out improvement direction.