计算机应用 ›› 2011, Vol. 31 ›› Issue (08): 2169-2173.DOI: 10.3724/SP.J.1087.2011.02169

• 信息安全 • 上一篇    下一篇

基于信息熵与动态区域分割的半脆弱数字水印算法

王海洋,车生兵,舒旭   

  1. 中南林业科技大学 计算机与信息工程学院,长沙410004
  • 收稿日期:2011-02-16 修回日期:2011-04-23 发布日期:2011-08-01 出版日期:2011-08-01
  • 通讯作者: 王海洋
  • 作者简介:王海洋(1984-),男,黑龙江拜泉人,硕士研究生,主要研究方向:信息隐藏、网络安全、数字图像处理;车生兵(1970-),男,湖南常德人,教授,主要研究方向:人工智能、网络安全、数字图像处理;舒旭(1984-),男,湖南溆浦人,硕士研究生,主要研究方向:信息隐藏、网络安全、数字图像处理。
  • 基金资助:

    中南林业科技大学研究生科技创新基金资助项目(2008sx14);湖南省研究生科研创新项目(CX2010B328)

Semi-fragile watermarking algorithm based on dynamic image segmentation and information entropy

Hai-yang WANG,Sheng-bing CHE,Xu SHU   

  1. College of Computer Science and Information Technology, Central South University of Forestry and Technology, Changsha Hunan 410004, China
  • Received:2011-02-16 Revised:2011-04-23 Online:2011-08-01 Published:2011-08-01
  • Contact: Hai-yang WANG
  • Supported by:

    Supported by the Graduate’s Scientific Research Fundation of Central South University of Forestry and Technology

摘要: 已有的半脆弱水印算法大多数采取了双步长的固定量化方法,对载体图像没有针对攻击特性给予考虑,只将载体图像分为平滑和纹理两个区域,使得水印的鲁棒性进入了一个瓶颈状态。为了进一步提高水印的鲁棒性,提出了一种基于图像动态分割技术和信息熵的半脆弱水印算法。图像动态分割技术将图像分为若干个嵌入区域,并且确定各个嵌入区域的嵌入强度;基于信息熵的步长量化算法将信息熵引入到步长量化的计算方法中,能够有效衡量不同嵌入区域所携带的敏感信息量。实验结果表明,提出的算法与已有半脆弱水印算法相比,视觉掩蔽性好,鲁棒性也有了较明显的提高。

关键词: 动态图像区域分割, 嵌入强度, 信息熵, 敏感信息, 半脆弱水印

Abstract: Most of the existing semi-fragile watermarking algorithms adopt the means of double-step fixed quantization, do not consider the attack characteristics for carrier image, and only divide the original image into smooth region and texture region, so that the robustness of watermarking has reached a bottleneck. To improve the robustness of watermarking further, the authors proposed a new semi-fragile watermarking algorithm based on the technique of dynamic image region segmentation and information entropy. The technique of dynamic image region segmentation divided an image into several embedding regions, and determined the strength of every embedding region; a quantization algorithm based on entropy introduced information entropy into the quantization algorithm and could effectively measure the amount of sensitive information carried by different embedding regions. The experimental results show that, compared with the existing semi-fragile watermarking algorithm, the proposed algorithm has better masking performance and stronger robustness.

Key words: dynamic image region segmentation, embedding strength, information entropy, sensitive information, semi-fragile watermarking

中图分类号: