计算机应用 ›› 2005, Vol. 25 ›› Issue (06): 1279-1281.DOI: 10.3724/SP.J.1087.2005.1279

• 信息安全 • 上一篇    下一篇

基于相关理论的小波域数字盲水印算法

边杏宾,朱清新   

  1. 电子科技大学计算机科学与工程学院
  • 发布日期:2011-04-06 出版日期:2005-06-01

Blind digital watermarking algorithm in wavelet domain based on correlation theory

BIAN Xing-bin, ZHU Qing-xin   

  1. Department of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu Sichuan 610054, China
  • Online:2011-04-06 Published:2005-06-01

摘要: 提出了一种新的图像数字盲水印算法。在水印嵌入过程中,用一个单元模板表示一个水印图像像素;在检测过程中,计算单元模板和嵌入水印后图像的小波变换系数的相关值,进行适当的门限处理就能提取出水印。水印检测不需要原始图像数据和水印位置信息,实现了水印图像的自同步盲检测。为了进一步提高该水印算法的鲁棒性和改善检测结果,在水印检测前使用拉普拉斯算子对图像进行锐化处理,使得所嵌入水印能抗击更多攻击种类,并使检测结果更清晰。在检测前加入锐化处理步骤,可以减小水印嵌入强度,而能得到同样的检测效果,因此也增强了水印的隐蔽性。实验表明,该算法具有较好的水印隐蔽性和鲁棒性。

关键词: 小波变换, 快速傅立叶变换, 相关理论, 锐化处理, 模板匹配

Abstract: A new blind digital watermarking algorithm was proposed. In embedding process, an unit module was used to represent a pixel, in detecting process correlation value of the unit module and watermarked image wavelet coefficients was calculated and was processed with proper threshold to detect the embedded watermark. The detecting process doesnt need original image and position information, so the scheme realizes blind detecting and self-synchronization. Sharpening the watermarked image with Laplacian before detecting could make the watermarking scheme more robust to attacks and get better detecting result. Furthermore, adding sharpening step could lower the demand of embedding strength without affect the detecting result. Thus it improved watermark invisibility. Experiments show that the algorithm has good invisibility and robustness.

Key words: wavelet transform, fast Fourier transform, correlation theory, sharpening processing, module matching

中图分类号: