数字水印攻击的一种新方法  

A New Type of Watermark Attack

作　　者：赵华[1] 袁盾韬[2] 林红军[3] 董春丽[3] 

机构地区：[1]河海大学计算机及信息工程学院 [2]中国南方电网超高压输电公司鲁布革发电总厂 [3]中国人民解放军信息工程大学

出　　处：《微计算机信息》2005年第07X期84-86,共3页Control & Automation

摘　　要：数字水印是数字产品产权保护的技术,是当前国际学术界研究的热点问题之一,它包括算法设计和攻击技术两方面,作为矛和盾的双方,既互相对立又相互促进,而当前的水印方案大都采用传统的攻击手段检验其鲁棒性,对攻击的研究较少。本文从理论上提出了一种基于盲信号分离技术的水印攻击方法,描述了其技术特点,并建立了相应的应用模型,可以推断,新方法的使用会对改进水印算法和水印检测的研究起到良好的作用。Much attention has been paid to digital watermark which is used to protect the property of data products .Technically, algorithm model and attack pattern are necessary in watermark designing,these two parts restrain and promot each other.Generally,conventional attack patterns are chosen to test watermark,however,it is turely helpful for watermark to experience with more attacks to gain the real robustness.This paper proposed a new approach to watermark attack which is based on the blind signal separation,we described its characteristics and put up a applicable model ,all these may have a good effect on the watermark studies henceforth .

关 键 词：数字水印 盲信号分离 水印攻击 鲁棒性 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]