论文部分内容阅读
根据序列中短重码的分布特征提出一种基于位平面相关性的隐写分析方法.文中定义了短重码间距统计概念,推导分析了统计分布特征.并由短重码间距统计可以反映序列相关性的特点,利用其统计结果在间距等于图像宽度处的统计优势,通过Poisson分布显著性检验检测隐秘消息.实验结果表明,在有效控制虚警概率的前提下,此算法对LSB(least significant bit)匹配隐写有良好的检测性能.另外此算法从位平面的相关特性着手进行隐写分析,在检测过程中不针对具体隐写算法,具有较强通用性.
A steganalysis method based on bit-plane correlation is proposed according to the distribution characteristics of short-weight codes in the sequence. The concept of short-weight code spacing is defined and the statistical distribution characteristics are deduced and analyzed. The short-distance codes can reflect the sequence correlation The statistical information is used to detect the covert messages by the significance test of Poisson distribution.Experimental results show that this algorithm is suitable for LSB (least significant bit) matching under the premise of effectively controlling the false alarm probability Steganalysis has a good detection performance.In addition, this algorithm starts with steganalysis based on the features of the bit plane, and does not target the specific steganalysis algorithm in the detection process, so it has strong universality.