J4 ›› 2014, Vol. 41 ›› Issue (1): 124-129.doi: 10.3969/j.issn.1001-2400.2014.01.022

• 研究论文 • 上一篇    下一篇

直方图平移的自适应大容量可逆水印算法

王祥1;李可1;付凯元2;骆璠1   

  1. (1. 西安电子科技大学 通信工程学院,陕西 西安  710071;
    2. 西安电子科技大学 计算机学院,陕西 西安  710071)
  • 收稿日期:2012-11-14 出版日期:2014-02-20 发布日期:2014-04-02
  • 通讯作者: 王祥
  • 作者简介:王祥(1983-),男,副教授,博士,E-mail: wangxiang@xidian.edu.cn.
  • 基金资助:

    国家自然科学基金资助项目(61202391,61172068);中国博士后科学基金资助项目(2012M521748);中央高校基本业务费资助项目(K5051201030,K50511010003);教育部新世纪优秀人才支持计划资助项目(NCET-11-0691)

Histogram shifting based adaptive reversible watermarking algorithm with a high capacity

WANG Xiang1;LI Ke1;FU Kaiyuan2;LUO Fan1   

  1. (1. School of Telecommunication Engineering, Xidian Univ., Xi'an  710071, China;
    2. School of Computer Science and Technology, Xidian Univ., Xi'an  710071, China)
  • Received:2012-11-14 Online:2014-02-20 Published:2014-04-02
  • Contact: WANG Xiang

摘要:

传统的基于直方图平移的可逆水印算法的单遍嵌入过程在每个像素点中至多只能嵌入1bit水印信息,为了达到更高的嵌入量必须使用多遍嵌入.但在第2遍以及其后的嵌入过程中,所计算的预测误差将受到嵌入的水印信息的影响,极大地降低嵌入效率.为此,笔者提出一种基于直方图可逆数字水印的改进算法,该算法根据图像的纹理复杂度在像素中自适应嵌入1bit或多比特水印信息,使得单遍嵌入量可以达到2bit/像素以上.此外,即使在单个像素点嵌入多比特水印信息,所使用的预测误差仍为原始预测误差,有效地降低了嵌入失真.实验结果表明,该算法在有效提高嵌入量的同时仍可保持较高的图像质量.

关键词: 自适应, 可逆水印, 直方图平移

Abstract:

The traditional histogram shifting based reversible watermarking algorithm can only embed at most 1.0 bit watermark information into one pixel in single-pass embedding. Therefore, we have to use multi-pass embedding to achieve a higher embedding capacity. However, in the second-pass and the subsequent embedding process, the prediction error will be affected by the embedding distortion, which reduces the embedding efficiency. To this end, this paper proposes an improved reversible watermarking algorithm by employing histogram shifting. The algorithm embeds 1.0 bit or more bit watermark in one pixel in terms of the context of image. Experimental results show that the algorithm can effectively improve the embedding capacity while maintaining a high image quality.

Key words: adaptive, reversible watermarking, histogram shifting

中图分类号: 

  • TN911.73
Baidu
map