Steganographic domain classification using multi-class

In this paper we propose a scheme to identify steganographic domains (including the pixel, DCT, and DWT domains). To this effect we analyze total 26 image quality measures summarized by Ismail Avcibas and choose some sensitive features based on the Analysis of Variance technique as feature sets to distinguish between cover-images and stego-images which are marked in pixel, DCT and DWT domain respectively. The classifier between cover and stego-images is built using multi-class Support Vector Machine on the selected quality metrics and is trained based on a Gaussian filtered version of the original image. The presented method can not only detect the presence of hidden message but also identify, the hiding domains. The experiment results show the proposed scheme achieved good classification results.