基于模糊聚类和HVS的DWT自适应盲水印算法研究  

A DWT-based Blind Image Adaptive Watermarking Scheme Utilizing HVS and Fuzzy Clustering Theory

作　　者：吴健珍[1] 谢剑英[1] 

机构地区：[1]上海交通大学电子信息工程学院自动化系,上海200030

出　　处：《计算机工程与应用》2004年第17期109-111,共3页Computer Engineering and Applications

摘　　要：文章提出了一种基于离散小波变换(DWT)的模糊自适应盲水印算法,根据人类视觉系统的亮度、纹理、以及对比度掩蔽特性,利用模糊聚类来决定水印的嵌入强度,使之可以自适应于图像,水印嵌入位置随机选择,并且水印的检测无需原始图像。为增强水印鲁棒性,笔者在图像的不同位置多次嵌入水印。通过仿真实验,表明算法可以在保证加水印后图像质量的前提下,以最大强度嵌入水印,对JPEG压缩、噪声、图像剪切具有较强的鲁棒性,由于水印检测无需原始图像,该方法具有实用性。This paper proposes a DWT based blind image adaptive watermarking scheme utilizing HVS and fuzzy clustering theory,through the classification of the image's brightness,texture and contrast sensitivity,people can embed more robust watermark randomly under a certain visual distance,in addition,watermark can be extracted without the access to the original image.In order to improve robustness,we embed the same watermark several times in different positions.Simulation results indicate that the scheme proposed in this paper can survive JPEG compression,noise as well as cropping,so it has the potential practical use.

关 键 词：数字水印 DWT HVS 模糊聚类 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]