基于SVM对隐写分析RS算法的改进

RS分析方法是隐写分析理论中检测LSB隐写的一种典型算法,但其对低密写率的情况下其正确检测率是不理想的.针对这种情况,结合统计学习理论,利用一种基于支持向量机(SVM)来改进RS隐写分析算法,在保留RS特征选取策略的前提下,改用支持向量机(SVM)对选取的特征集进行分类识别.实验结果表明,原始无损存储图像,经改进后的算法比RS隐写分析算法具有更优的性能.

A novel RS steganalysis algorithm based on support vector machines

RS steganalysis method is a typical arithmetic in steganalysis theory used for testing steganography least significant bit, but its accuracy is not ideal in the case of the circumstance of low embedded rate. As a response to this problem, a steganalysis algorithm based on statistis theory and support vector machines is proposed in this paper. The revised arithmetic uses SVM to classify selected characteristic set on the condition that RS characteristic selected strategy is kept. The experiment shows that the revised arithmetic is superior to RS steganalysis algorithm in original raw losslesslv stored images.