Steganalytic method based on short and repeated sequence distance statistics  

作　　者：WANG GuoXin PING XiJian XU ManKun ZHANG Tao BAO XiRui 

机构地区：[1]Institute of Information Engineering, Information Engineering University, Zhengzhou 450002, China

出　　处：《Science in China(Series F)》2008年第10期1466-1474,共9页中国科学（F辑英文版）

基　　金：the National Natural Science Foundation of China (Grant No.60473022)

摘　　要：According to the distribution characteristics of short and repeated sequence （SRS）, a steganaiytic method based on the correlation of image bit planes is proposed. Firstly, we provide the conception of SRS distance statistics and deduce its statistical distribution. Because the SRS distance statistics can effectively reflect the correlation of the sequence, SRS has statistical features when the image bit plane sequence equals the image width. Using this characteristic, the steganalytic method is fulfilled by the distinct test of Poisson distribution. Experimental results show a good performance for detecting LSB matching steganographic method in still images. By the way, the proposed method is not designed for specific steganographic algorithms and has good generality.According to the distribution characteristics of short and repeated sequence （SRS）, a steganaiytic method based on the correlation of image bit planes is proposed. Firstly, we provide the conception of SRS distance statistics and deduce its statistical distribution. Because the SRS distance statistics can effectively reflect the correlation of the sequence, SRS has statistical features when the image bit plane sequence equals the image width. Using this characteristic, the steganalytic method is fulfilled by the distinct test of Poisson distribution. Experimental results show a good performance for detecting LSB matching steganographic method in still images. By the way, the proposed method is not designed for specific steganographic algorithms and has good generality.

关 键 词：information hiding STEGANALYSIS statistics of short and repeated sequence distance 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]