基于图像LSB的隐藏检测算法研究

摘要

This paper researches on three algorithms of RS,QRS,and IRS（only R）,detecting LSB steganography in images and presents an improved algorithm based on the QRS and IRS（only R） algorithms,called QRS（only R）.Simulation results of a large number of images show that： compared to the RS,the QRS and the IRS（only R）,this new algorithm has improved performances on the detection rate,the accuracy and the executive time,and their improved relative values can be up to 17%,30% and 38%.The algorithm can detect all kinds of LSB steganography algorithms reliably and is a simple and practical steganalysis algorithm.%文章研究了RS（Regular Groups and Singular Groups）、QRS（Quadratic RS Steganalysis）和IRS（仅R）（Improved RS by OnlyRegular Groups）三种图像隐藏检测算法,并提出了一种将QRS和IRS（仅R）算法思想相结合的改进算法,即QRS（仅R）。通过大量图像的仿真表明：该算法相比于RS、QRS、IRS（仅R）算法,在检测率、估计精度和检测时间上均有所改善,改善的相对值分别高达17%、30%和38%。该算法对各种LSB（Least Signification Bit）隐藏算法均能有效检测,是一种简单实用的隐写分析算法。