计算机与现代化

• 图像处理 • 上一篇    下一篇

DCT域内视频分形脆弱水印算法

  

  1. (1.青岛科技大学数理学院,山东青岛266061; 2.亚马逊A9中央经济搜索部,美国硅谷94301; 3.中国石油大学(华东)石油工程学院,山东青岛266580)
  • 收稿日期:2017-04-02 出版日期:2017-12-25 发布日期:2017-12-26
  • 作者简介:杨树国(1970-),男,山东曹县人,青岛科技大学数理学院教授,硕士生导师,博士后,研究方向:数字水印,应用数学,数学建模; 张波(1995-),男,河南开封人,本科生,研究方向:数学建模,数字水印。
  • 基金资助:
    山东省重点研发计划项目(2015GGX101020); 山东省研究生教育创新计划项目(SDYY16010); 山东省教育科学“十二五”规划课题(YBS15014); 山东省研究生导师指导能力提升项目(SDYY17042)

A Fragile Video Watermarking Algorithm Based on Fractal Theory

  1. (1. School of Mathematics & Physics, Qingdao University of Science & Technology, Qingdao 266061, China;  2. Amazon A9, 101 Lytton Ave, Palo Alto, CA 94301, USA; 3. School of Petroleum Engineering, China University of Petroleum, Qingdao 266580, China)
  • Received:2017-04-02 Online:2017-12-25 Published:2017-12-26

摘要: 针对视频完整性认证问题,提出一种DCT域内视频分形脆弱水印算法。首先把视频I帧的RGB色彩模式转换为YUV色彩模式,并对Y分量进行DWT变换;然后提取I帧的分形特征,结合每一帧的位置特征生成认证码,用Logistic混沌映射将认证码置乱;最后将置乱的认证码嵌入每一帧DCT块最后2个非0系数的LSB位上。在视频认证过程中,无需参考原视频,实现了盲检测。实验结果表明,该方法在保证原视频质量的情况下,能有效检测到对原视频的篡改和攻击,并能定位篡改和攻击的位置。

关键词: 脆弱水印, 离散小波变换, 离散余弦变换, 分形

Abstract: Aiming at the video integrity authentication problem, this paper proposes a fragile video watermarking algorithm based on fractal theory, which converts the color pattern RGB into color pattern YUV and embeds the watermark into the component Y. Transformed by DWT, the fractal feature of I-frame is extracted, which combines with the position feature of each frame to generate authentication code. Then, the authentication code which is scrambled by Logistic chaos mapping is embedded into the LSB bit of last two non-zero coefficients of DCT block. In the process of video authentication, it is possible to realize the blind detection without reference to the original video. The experimental results show that the proposed algorithm can effectively detect and locate the position of the original video which is attacked and tampered with, without degrading image quality.

Key words: fragile watermarking, DWT, DCT, fractal

中图分类号: