摘要
传统的JPEG图像盲隐写检测算法主要是通过单个或是两个特征集融合的方式来设计。而对于多个特征集,如何从这些特征集中选取一个较优的组合进行融合,目前尚处于研究阶段。本文提出一种基于改进AdaBoostSVM的多特征融合隐写检测方法,并通过设计多个实验来验证算法的性能。通过实验比较,在JPEG图像隐写多特征融合盲检测中,该方法能够在有限的计算复杂度下得到一个融合效果较优的特征组合。
Traditional blind steganalysis algorithm for JPEG image is mainly designed by a single or two feature-sets fusion.However,for multiple feature sets,how to select an optimum combination to fuse from these feature sets is still at a research stage.This paper proposes a universal steganalysis method for multi-feature fusion based on the improved AdaBoostSVM JPEG image,and many experiments are designed to verify the performance of the algorithm.By experimental comparison,it is found that this method can obtain a feature combination of optimum fusion effects with limited computational complexity in JPEG-imaged blind steganalysis for multi-feature fusion.
出处
《计算机与现代化》
2013年第7期120-122,238,共4页
Computer and Modernization
基金
福建省科技创新平台项目(2009J1007)
福建省自然科学基金资助项目(2010J01331)
