摘要
作为版权保护的重要手段,数字水印技术已得到了广泛的研究和应用,但是实用的彩色图像数字水印技术不多,特别是多通道彩色图像数字水印技术。本文首先从原始彩色图像中提取红色分量、绿色分量和蓝色分量,利用DWT变换对各分量分别进行多级小波分解,然后在各分量的第L级的水平、垂直、对角线小波分解系数中自适应地重复地嵌入水印,而在水印信息检测时,多次采用了“多数原则”,并且不需要原始图像。经实验测试,本文提出的算法对诸如JPEG压缩、滤波、噪声污染、亚抽样等常见的信号处理和攻击均具有较强的鲁棒性。
As a main method for copyright protection, the techniques of watermarking have been widey,esearched and applied, but there are few methods of practical color image digital watermarking, especially in the case of multichannel color image digital watermarking. In this paper, red component, green component and blue component are first extracted from the raw color image and decomposed using DWT in ultiresolution,then encrypted bilevel watermarking is adaptively repeatedly embedded into the horizontal, vertical and diagonal wavelet coefficients of multiresolution representation of each component in the L layer, and receive a good tradeoff between invisibility and robustness, detecting procedure need not origin host image by “majority principles” repeatedly. The experimental results show that the proposed method is effectively robust against JPEG compression, commonly used image processing and attack such as filtering and noise and sub-sampling etc.
出处
《现代计算机》
2005年第12期40-42,91,共4页
Modern Computer
