A Novel LSB Matching Steganalysis Detection Algorithm Based on Characteristics of the Second Order Difference Markov

For the LSB matching steganography with high security, the author has calculated the second-order differential of the image pixels on horizontal and vertical directions, obtaining the differential matrix and taking it as sensitive feature extracting source. Then, by extracting the second-order Markov transformation matrix of the differential matrix as the features, this paper proposed a LSB matching steganalytic algorithm. Experimental results show that: compared with existing algorithm, the proposed algorithm can not only preserve the high detection accuracy, but also speedup the detection process in a large extent, which enhance the performance and practicability of the steganalytic algorithm.