Universal steganalysis to images from normal and forged images

We proposed a CPMF (Combinative probability mass function) model for universal steganalysis to images from nature images and forged images, which can present statistic divergence between cover or forged image and steg image prominently. Based on the two features of images and CPMF mode, we discussed the affection of steg process and forged process on the nature images. This model relied most on the statistic relativity of between two color channels. At last, many experiment results are given to support our theory.

[1]  Xiongfei He,et al.  Classification between PS and Stego Images Based on Noise Model , 2009, 2009 Third International Conference on Multimedia and Ubiquitous Engineering.

[2]  Alin C. Popescu,et al.  Exposing digital forgeries in color filter array interpolated images , 2005, IEEE Transactions on Signal Processing.

[3]  Siwei Lyu,et al.  Steganalysis using color wavelet statistics and one-class support vector machines , 2004, IS&T/SPIE Electronic Imaging.

[4]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[5]  B. S. Manjunath,et al.  Steganalysis for Markov cover data with applications to images , 2006, IEEE Transactions on Information Forensics and Security.

[6]  Shih-Fu Chang,et al.  Blind Detection of Digital Photomontage using Higher Order Statistics , 2004 .

[7]  Siwei Lyu,et al.  Detecting Hidden Messages Using Higher-Order Statistics and Support Vector Machines , 2002, Information Hiding.