北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2017, Vol. 40 ›› Issue (s1): 34-38.doi: 10.13190/j.jbupt.2017.s.008

• 论文 • 上一篇    下一篇

一种新的结合SVD的非下采样Shearlet数字水印

赵健1, 张婉如1, 张顺利1, 雷敏2, 徐文胜1, 范帅帅1   

  1. 1. 西北大学 信息科学与技术学院, 西安 710127;
    2. 北京邮电大学 信息安全中心, 北京 100876
  • 收稿日期:2016-04-14 出版日期:2017-09-28 发布日期:2017-09-28
  • 作者简介:赵健(1973-),男,教授,硕士生导师,E-mail:zjctec@nwu.edu.cn.
  • 基金资助:
    国家自然科学基金项目(61572400,61379010);陕西省自然科学基础研究计划项目(2015JM6293)

A Novel Hybrid Digital Watermarking Scheme Using SVD in Non-Subsampled Shearlet

ZHAO Jian1, ZHANG Wan-ru1, ZHANG Shun-li1, LEI Min2, XU Wen-sheng1, FAN Shuai-shuai1   

  1. 1. School of Information Science and Technology, Northwest University, Xi'an 710127, China;
    2. Information Security Center, Beijing University of Posts and Telecommunications, Beijing 100876, China
  • Received:2016-04-14 Online:2017-09-28 Published:2017-09-28

摘要: 提出了一种非下采样剪切波变换域下的分块数字水印算法,充分利用了非下采样剪切波变换中的方向性特征,并选用信息熵作为选择图像中待嵌入块的标准,应用奇异值分解的方法进行水印信息的嵌入,解决了多尺度分析水印算法中不可见性和鲁棒性不能兼顾的问题. 实验结果表明,提出算法中峰值信噪比高于51 dB,归一化相关系数均高于0.93,其性能得到了较大的提高.

关键词: 数字水印, 非下采样剪切波变换, 方向性特征, 奇异值分解, 信息熵

Abstract: A novel block-based digital watermarking scheme in non-subsampled shearlet transform was proposed. The directional features of the NSST was utilized and the entropy was taken as the criterion of selecting the embedded blocks in images. Besides, this scheme applied singular value decomposition to embed the binary watermark logo. The algorithm not only solved the problem that the invisibility and the robustness in multiscale analysis watermarking algorithm could not be taken into account, but also fully showed the superiority comparing with other previous schemes. Experiment results demonstrated that the peak signal to noise ratio was above 51 dB and most of normalized correlation values were higher than 0.93. The imperceptibility and robustness performance was experimentally proved to be enhanced.

Key words: digital watermark, non-subsampled shearlet transform, directional features, singular value decomposition, information entropy

中图分类号: