RDWT和NVF的自适应彩色图像水印算法

提出一种基于冗余小波(RDWT:Redundant Wavelet Transform)和噪声可见度函数(NVF:Noise Visibility Function)的彩色图像水印嵌入新算法,该算法利用主体图像的稳态广义Gaussian模型描述图像的局部特性,建立了水印信息的加权系数计算公式。在水印嵌入过程中先对主图像在YIQ空间的Y分量进行RDWT变换。同时利用NVF函数对经过加密变换的水印进行加权处理,然后利用奇异值分解将水印嵌入到变换域复杂纹理和边缘区域。实验结果表明该算法在保证水印不可见性的同时能很好地抵抗多种几何攻击和图像处理攻击,实现了水印鲁棒性、不可见性和容量的统一。

The Algorithm of Content Adaptive Color Image Watermark Based on RDWT and NVF

In this paper,a content adaptive color image watermarking algorithm based on the RDWT and NVF is presented.This algorithm is based on a stationary generalized Gassian model with local image properties,and watermark is embeded into the texture and edge region for more strongly embedded watermark by noise visibility function.The suggested method is performed by modification on Singular Value Decomposition(SVD) of images in RDWT of Y component in YIQ color space.Experimental results show that the algorithm keep the balance of robustness,trans-parence and capacity.The algorithm is adaptive to all sorts of meaningful watermarks.