盲图像水印方法仿真研究

Simulation of Blind Watermarking Method

研究水印的图像版权保护问题,由于要求证明水印的存在和代表所有权,针对图像水印加载算法容易导致原始图像信息丢失的缺陷和破坏保护特性,提出了一种小波树形结构的量化算法,使得在水印的提取过程中,无需原始图像或水印信息即可还原出无损的原始图像。首先利用四级小波变换,构造出树形小波系数波段图。对树型结构中不同波段节点,可依据系数差异最大原则进行离散化并嵌入水印信息。通过在提取过程中设计自适应阀值,可以更好地抑制一些常见的水印损失,如滤波、均值过滤和高斯噪声等。实验结果表明,改进方法能使水印图像在视觉上与原始图像一样,起到保护版权的有效性。

The implementation of watermarking for a blind image using wavelet tree quantization for copyright protection is proposed.Aiming at image watermark algorism which easily results in the loss of information of initial image,a quantization algorism based on wavelet tree structure is introduced,which makes it not need any original image or watermark during the watermark extraction.Firstly,wavelet coefficients band graph is constructed by using tree wavelet four-level wavelet transform;then according the maximum coefficient difference principle,different band node of the tree-structure is discrete and embedded with the watermark.Moreover,the common attacks,such as median filtering,average filtering,and Gaussian noise,can be resisted by designing an adaptive threshold value in the extraction process.The experiment shows the watermarked image looks visually identical to the original,and the watermark can be effectively extracted even after image pre-processing.