基于图像归一化的轮廓域数字水印算法研究

Study on Contourlet-domain digital watermarking algorithmbased on image normalization

为了使Contourlet域数字水印能有效地抵抗几何攻击,提高数字水印的鲁棒性,以图像归一化理论为基础,提出了一种可有效抵抗几何攻击的Contourlet域数字图像水印方案。该方案利用基于矩的图像归一化技术和不变质心理论提取出归一化图像的重要区域,对该重要区域进行Contourlet变换,根据Contourlet变换后图像系数的特性,用奇偶量化的方法将水印信息嵌入到Contourlet变换后中频子带上。实验结果表明,该方案不仅具有很好的透明性,而且具有较强的抵抗常规信号处理及几何攻击的能力。

In order to improve the resistance of digital watermarking in Contourlet domain to geometric attacks and to improve the robustness of digital watermarking,a watermarking Contourlet scheme for digital image that can resist geometrical attacks is proposed. Firstly,the significant region of normalized image is extracted with Zernike moment-based image normalization and invariant centered theory.Then,Contourlet transformis performed on the significant region.Finally,by making the best use of the coefficients characteristics of the image after Contourlet transform,the meaningful watermark is embedded into the middle frequency band after Contourlet transform with odd-even quantization method.The experimental results show that the scheme has a good transparency and robust to both geometric attacks and signal processing attacks.