基于DWT的数字图像水印算法研究

研究了一种基于小波变换(DWT)的自适应图像水印算法。该算法引入了基于噪声可见函数(NVF)的感知模型来描述图像的局部特征,并根据此模型决定水印信息的嵌入深度。水印嵌入过程是:首先对载体图像进行小波分解并选择中频部分做为水印的嵌入区域;然后通过计算中频部分的噪声可见函数决定嵌入因子;最后通过小波反变换得到嵌入水印的图像。实验证明,该算法对具有不同特性的灰度图像的自适应性较强,且对常见的攻击具有较强的鲁棒性。

Research of Digital Image Watermarking Algorithm Based on DWT

An adaptive image watermarking algorithm based on DWT is researched.The algorithm introduces the perceptual model based on NVF（noise visibility function） to describe local characteristics of images,and the embedding depth is determined by the model.The watermarking embedding process is as follows： firstly the carrier image is decomposed by DWT and the mid-frequency section is selected as watermarking embedding area;then the embedding factor is determined through computing NVF of the mid-frequency section;finally the watermarked image is obtained by reverse DWT.The experiments show that the algorithm has strong self-adaptability to various gray images with different characteristics and robustness to common attacks.