基于DWT具有篡改定位的半脆弱水印算法  

Semi-Fragile Watermarking Algorithm Based on DWT with Tamper Localization

作　　者：李子臣 宋培非 丁海洋 Ll Ziehen;Song Peifei;Ding Haiyang(School of Information Engineering,Beijing Institute of Graphic Communication,Beijing 102600)

机构地区：[1]北京印刷学院信息工程学院,北京102600

出　　处：《信息安全研究》2019年第2期162-167,共6页Journal of Information Security Research

基　　金：国家自然科学基金项目(61370188);北京市教委科研计划一般项目(KM201610015002;KM201510015009);北京市教委科研计划重点项目(KZ201510015;KZ201710015010);科技创新服务能力建设科研水平提高定额项目(PXM2017_014223_000063);北京印刷学院校级资助项目(Ec201803 Ed201802 Ea201806)

摘　　要：半脆弱水印因为在多媒体内容认证方面的重要作用而受到人们密切的关注.为了能够区分偶然攻击与恶意篡改,半脆弱水印需要对一般的内容保护图像操作有一定的鲁棒性.根据图像相邻小波高频系数之间的大小关系在JPEG压缩之后大多数没有发生变化这一事实,提出了一种新的基于DWT变换的具有篡改定位的半脆弱水印算法,该算法主要思想是根据图像相邻小波高频系数之间的大小关系来嵌入水印.仿真实验结果表明:该算法嵌入容量大,有很好的抗JPEG压缩性能,同时也能对恶意篡改精确定位.Semi-fragile watermarking has attracted much attention because of its important role in multimedia content authentic.ation. In order to distinguish accidental attacks and malicious tamperingi semi-fragile watermarks must have the robustness to process general content protection and image manipulation. After JPEG compression, most of the size relationships between the highfrequency coefficients of adjacent wavelets of an image have no changes. According to this fact,this paper proposes a new semi-fragile watermarking algorithm based on DWT transform with tamper localization. The watermark is embedded according to the size relationship between the high-frequency coefficients of the adiccent wavelets of the image. The results of simulation experiments show that the algorithm has large embedded capacity and good anti-JPEG compression performance. At the same time, the positioning for malicious tampering is also very accurate.

关 键 词：半脆弱水印 恶意篡改 JPEG压缩 小波系数 高频系数 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]