计算机工程

• 图形图像处理 • 上一篇    下一篇

基于异常点检测的通用无监督隐写取证算法

吴运达,张涛,侯晓丹,徐琛   

  1. (解放军信息工程大学 信息系统工程学院,郑州 450000)
  • 收稿日期:2016-10-31 出版日期:2017-11-15 发布日期:2017-11-15
  • 作者简介:吴运达(1992—),男,硕士研究生,主研方向为图像处理、信息安全;张涛,教授、博士;侯晓丹,博士;徐琛,硕士。
  • 基金项目:
    国家自然科学基金(61572518,61272490)。

Universal Unsupervised Steganalysis Forensics Algorithm Based on Outlier Detection

WU Yunda,ZHANG Tao,HOU Xiaodan,XU Chen   

  1. (Information System Engineering Constitute,PLA Information Engineering University,Zhengzhou 450000,China)
  • Received:2016-10-31 Online:2017-11-15 Published:2017-11-15

摘要: 针对图像隐写分析时存在的载体来源失配问题,提出一种结合图像检索和异常点检测的通用无监督隐写取证算法。对待测图像,从图像数据集中检索具有相似统计特性的载体图像构造辅助图像集,而载密图像可视为载体图像中的异常点,通过异常检测的方法实现载密图像的无监督通用盲检测,避免失配问题和复杂的分类器设计过程。对于低维检测特征和富模型特征,分别使用针对基于密度和基于高维空间的异常检测算法进行隐写检测。实验结果表明,与典型的空域隐写算法相比,该算法具有更高的检测效率。

关键词: 图像隐写分析, 通用盲检测, 图像滤波检测, 图像检索, 局部异常因子, 异常点检测

Abstract: In order to solve the carrier source mismatch problem in image steganalysis,a universal unsupervised steganalysis forensics algorithm combining image retrieval and outlier detection is proposed.For the measured image,The carrier images with the same statistical properties are retrieved from the image database as aided samples.The dense image can be regarded as the outlier in the carrier images.The unsupervised general blind detection of dense images is realized by means of anomaly detection,which avoids mismatch problems and complex classifier design process.The density based and high dimensional space based anomaly detection methods are used respectively for low dimensional detection features and rich model features.Experimental results show that compared with the typical spatial steganography algorithms,the proposed algorithm has better detection efficiency.

Key words: image steganalyasis, universal blind detection, image filtering detection, image retrieval, Local Outlier Factor(LOF), outlier detection

中图分类号: