小波域内JND数字水印算法的改进

提出了一种鲁棒型盲数字水印算法.在修改基于人类视觉模型的JND阈值基础上,利用契比雪夫多项式迭代所产生的混沌序列作为图像的数字水印,提出了一个新颖的盲水印嵌入与检测算法.实验结果表明,该算法对于常见的图像变换,例如中值滤波、旋转、平移、剪切、JPEG压缩、加噪等都具有较强的鲁棒性,特别是对含有水印的图像在所有角度上的旋转、任意的平移及剪切时,仍能够检测到图像中的水印.

An improved just noticeable differences watermarking algorithm on wavelet domain

A robust blind digital watermarking algorithm was developed by modifying the threshold of just noticeable differences(JND) based on a human visual system(HVS) and applying the iterations of a Chebyshev polynomial to generate a chaotic sequence as digital watermarking of an image.Experiments show that the algorithm has strong robustness for typical image transforms such as median filtering,rotating,shifting,cropping,JPEG compression,and noise adding.Furthermore,the watermark can also be detected when the watermarked image processes all angle rotating or arbitrary shifting or cropping.