Steganalysis Versus Splicing Detection

摘要: Aiming at detecting secret information hidden in a given image using steganographic tools, steganalysis has been of interest for years. In particular, universal steganalysis, not limited to attacking specific tool, is extensive interests due its practicality. Recently, splicing detection, another important area digital forensics attracted increasing attention. Is there any relationship between and detection? it possible apply methodologies this paper, we address these intact yet interesting questions. Our analysis experiments have demonstrated that, on the one hand, steganography different goals strategies, hence, generally causing statistical artifacts images. However, other both them make touched (stego or spliced) from corresponding original (natural) image. Therefore, natural model based set carefully selected features under machine learning framework can be used detection. It shown paper that some successful steganalytic schemes promising progress detection if applied properly. A more advanced developed state-of-the-art methods thereafter presented. Furthermore, concrete implementation proposed Columbia Image Splicing Detection Evaluation Dataset, which achieved an accuracy 92%, indicating significant advancement