A DCT Steganographic Classifier Based on Compressive Sensing
Constantinos Patsakis, Nikolaos Aroukatos · 2011
Due to DRM, steganographic techniques have received a lot of focus and more sophisticated techniques are continuously being developed. As a mean to estimate their strength in data hiding, steganalysis has recently received a lot of focus from researchers. This work, addresses to the problem of identifying a clean image from a set where other instances of the same image exist, but data have been embedded in DCT. A new effective steganographic classifier is presented which has very good properties. The novelty of the proposed method is the use of compressive sensing, that seems to have big impact on steganalysis.