一种可估计最低位平面嵌入信息量的密写分析方法

A Steganalytic Method Capable of Estimating LSB-Embedding Capacity

检测图像中有无隐蔽的密写数据是信息安全的重要课题,当前一种常用的密写技术是将秘密数据嵌入图像最低位面,称为LSB密写.分析表明,LSB密写对载体图像统计特性的影响能用来估计嵌入的信息量,由此给出一种有效的密写分析新方法.根据待检测图像的直方图特性定义2个参数,并对待检测图像进行二次密写,观察这2个参数的变化趋势,从而估计出嵌入的数据量.该方法采用二次密写,不仅能判断密写信息的存在性,而且给出隐藏的数据量,具有实现简单、计算量小的优点.实验表明,估计出的密写信息量比较可靠.

Detection of hidden message in still images is an important research topic in information security. A popular method to embed data in an image is to modify the least significant bit(LSB) plane of the host image. It is shown that the statistical changes induced by data embedding can be used to estimate the embedding capacity, leading to a new staganographic technique. Two parameters are defined based upon histogram of the image under investigation, and a test embedding is performed to observe the change in these parameters. This gives an estimate of possible stego-data. The proposed method uses a secondary embedding, which not only reveals the presence of a secret message, but also provides an estimate of the payload. The algorithm can easily be implemented with low computational complexity. The experimental results show that the estimated payload is quite reliable.