基于神经网络自联想的数字图像水印算法

Neural Network-Based Digital Image Watermarking Algorithm

基于神经网络自联想记忆功能,提出了一种新的数字图像水印算法.首先将水印信息存储到神经网络中;再将其分别嵌入到原始图像小波分解的近似系数分量和细节系数分量中,采用嵌入水印过程的逆过程提取水印;最后利用神经网络对提取出的水印图像进行后处理.实验结果表明所提方法在满足水印的不可感知性的前提下,对多种攻击,如JPEG压缩、加性和乘性噪声干扰、滤波、剪切等具有较强的鲁棒性.

A novel digital image watermarking algorithm based on the association memory ability of neural network is proposed in this paper.The watermark is stored in a neural network firstly;then embed into the significant and detail wavelet coefficients of original image.The inverse embedding process is first used to extract watermark,then the extracted watermark is processed using the association memory ability of neural network.Experimental results show that the algorithm can improve the watermark image quality which is extracted from the watermarked image attacked by JPEG compression,noise adding,filtering,cropping etc.