Steganalysis by Tracking Image Noise with Higher Order Statistics

摘要

We present statistical techniques for Steganalysis of natural images potentially subjected to data hiding. Our hypothesis is that image noises have certain statistics, while data hiding schemes will alter them and generate entirely different noises that can be tracked with high order statistics. We extract noises from images by wavelet denoising which outperforms Gaussian smoothing with respect to images with good quality. In this paper, wavelet packets decomposition followed with band selection is adopted to obtain statistics of image noises instead of wavelet decomposition. The classifier between cover-images and stego-images is built using support vector machines (SVM) which have good generalization performance.