应用提升小波变换的均值量化数字水印方法

Novel Method of Mean Quantization Digital Watermarking Based on Lifting Wavelet Transform

针对数字图象的版权问题,提出一种基于提升小波变换的均值量化数字水印方法.该方法将提升小波变换和均值量化相结合,首先对水印图象进行Arnold变换将其置乱,然后将原图象的像素矩阵划分为若干小块,并进行二维提升小波变换,接着对小波的低频系数进行均值量化处理,实现其水印嵌入与盲提取.实验结果表明,该方法不仅实现了水印的盲检测,而且对常见的攻击手段如剪切攻击、噪声攻击等具有很强的适应性,即使经历多重攻击依然能够准确地提取出水印图象.

To the copyright issues of digital image,a novel method of mean quantization digital w atermarking based on lifting w avelet transform is presented. This method combines lifting w avelet transform and mean quantization,it firstly scrambles the w atermarking image through Arnold transform,then it divides the pel matrix of original image into small pieces and takes lifting w avelet transform.After that,it takes mean quantization process on the low frequency coefficient of w avelet to achieve the embedding and blind extraction of w atermarking. The experimental results show that this method not only achieve the blind extraction of w atermarking,but also it has a good adaptability on common attack means,such as shear attack,noise attack etc,and it can extract the w atermarking image accurately even after multiple attacks.