自适应阈值小波滤波复原在隐写分析中的应用

Application of adaptive thresholding wavelet filter restoration in steganalysis

将空域图像的数字隐写过程看作是一种被加性噪声污染的图像退化过程,从而引入图像复原技术来进行秘密消息的检测。将自适应阈值小波滤波应用于隐写分析,并结合图像滤波前后的差异性特征提出一种新的LSB匹配隐写分析方法。实验结果表明,该方法要优于Ker的校准法,尤其对未压缩的高精度原始图像的检测更为有效,并且不用估计嵌入率,方法简单,易于实现。

When the steganography in spatial domain was modeled as an image degradation process caused by additive noise, the image restoration could be introduced for secret messages detecting. A new method of LSB matching steganalysis was proposed through using adaptive thresholding wavelet filter restoration in steganalysis and combine with the difference characters between the original image and the filtered one. Experimental results show that, the method is better than Ker＇s calibrating method, especially in detecting uncompressed high resolution raw image due to its simple realization without estimating the embedding rate beforehand.