信息安全研究 ›› 2016, Vol. 2 ›› Issue (6): 490-500.

• 学术论文 • 上一篇    下一篇

网络大数据下的信息隐藏分析

孔祥维   

  1. 大连理工大学
  • 收稿日期:2016-06-14 出版日期:2016-06-15 发布日期:2016-06-15
  • 通讯作者: 孔祥维
  • 作者简介:教授,博士生导师, 主要研究方向为多媒体信息安全、数字图像处理和识别、大数据下的多媒体语义理解、多媒体知识管理和商务智能、多源信息感知和信息融合等.

Steganalysis of the Big Data over Network

  • Received:2016-06-14 Online:2016-06-15 Published:2016-06-15

摘要: 随着网络信息爆炸式的增长,近年来信息安全引起了广泛的关注.信息隐藏作为信息安全领域的热门方向,也同样面临着大数据带来的挑战.网络中存在着海量的数据且这些数据来源多样、质量不一等诸多特点,使很多实验室环境下的信息隐藏分析方法因此失效.以数字图像为例:首先分析了网络大数据下数字图像的特点.然后阐述了实验室环境下信息隐藏分析方法的国内外研究成果.通过实验说明了实验室环境下信息隐藏分析方法的优异性能在网络大数据环境下急剧下降,重点从对数据集研究、对特征的处理、对分类器的改进以及整体结构的改变4个方面介绍了当前针对网络大数据下存在的信息隐藏分析失配问题进行的研究工作.最后指出了网络大数据下信息隐藏分析的潜在问题与未来挑战.为解决大数据下信息隐藏分析问题提供了有效的方法借鉴.

关键词: 信息隐藏分析, 大数据, 网络, 大数据安全, 比较研究

Abstract: With the explosive growth of information on the Internet, information security has caused wide public concern over the recent years. Steganalysis as a popular direction in the field of information security, is also facing the challenges posed by big data. The data over network always presents in huge amount and comes from multiple sources and different qualities, which makes the steganalysis methods under laboratory become invalid. This paper takes digital images as an example, at first introduces the research situation of steganalysis under laboratory environment, and then analyzes the characteristics of image under the big data over network. Steganalysis methods under laboratory always have dramatic performance decrease under the big data over network by experiment. The paper focuses on the research on the mismatch of steganalysis of the big data over networks, researches on the key technology and the latest progress form four aspects which are the research on the data set, the process on the feature, the improvement of the classifier and the change of steganalysis framework . In the end, the potential problems and future challenges of steganalysis of the big data over network are pointed out, which provide a relatively comprehensive reference basis for future research.

Key words: steganalysis, big data, network, big data security, comparison study