基于AVC和小波变换的数字图像水印嵌入量的研究

鲁棒性、不可见性和水印嵌入量是评价图像数字水印质量的3个重要指标，而水印嵌入量又是嵌入算法质量好坏的关键，在分析利用传统通信信道计算水印嵌入量的缺陷和不足的基础上，提出了一种新的计算水印嵌入量算法．该算法不仅分析了图像数字水印的特性，结合了视觉模型和小波变换的特点，而且利用了可变信道（AVC）的原理。实验结果证明，该算法不仅水印嵌入量大，而且水印的鲁棒性和不见性也相当高，此算法的出现为进一步实现信息的隐藏奠定了理论基础。

Watermark capacity of image based on AVC and wavelet transform

For image watermark,robustness,imperceptibility and capacity are three important factors that affect the watermark quality.The most important is capacity that controls the other factors.So the purpose of the paper is to find the theoretical capacity bound of digital image watermark based on human vision system(HVS) model and wavelet transform.First,the limitations of traditional algorithms that apply the ordinary communicational channel on the image watermark capacity is showed.With the characters of wavelet transform,a new algorithm based on HVS models and arbitrarily varying channel(AVC) is proposed,which not only can ensure the watermark information robust and imperceptibility,but also can accord with the characters of image channel.