A DCT Statistics-Based Universal Image Steganalysis

Universal steganalysis detect the presence of secret messages. In this paper, we proposed a new methodology which will be created different groups of images based on image features as an unsupervised learning and proposed methodology can enable us to design specific models of steganalyzer in order to improve system accuracy. These models will be differentiated all the infected JPEG images from the rest of JPEG images as a supervised learning. In the current steganalysis systems, all images with different statistical properties are treated equally but in our method, we can distinguish between images to build better steganalyzers. For this purpose, we have used fuzzy clustering algorithm to construct images models specific in the training phase and have also used fuzzy decision making strategy to predict the class label of test data in the comparison phase. The above features focused on DCT statistics and could be used to distinguish between images. We have employed Large Scale Support Vector Machines (SVMs) to design our models. Experiments confirm the advantage of the proposed method over the current steganalysis systems.