基于SIFT和Radon变换的盲水印算法  

Blind watermarking algorithm based on SIFT and Radon transform

作　　者：王玉成 陈青[1] WANG Yucheng;CHEN Qing(School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)

机构地区：[1]上海理工大学光电信息与计算机学院,上海200093

出　　处：《电子设计工程》2022年第21期180-184,共5页Electronic Design Engineering

基　　金：上海理工大学国家级项目培育基金(16HJPY-MS06)。

摘　　要：针对当前数字水印算法水印的嵌入区域固定而易遭破坏且抵抗旋转攻击能力弱的问题,提出一种基于Radon变换和SIFT特征提取的盲水印算法。采用SIFT算法对图像进行特征提取,结合RANSAC算法筛选出稳定特征点,根据特征点选取合适大小的水印嵌入区域,经过DWT和SVD分解将水印信息嵌入在选定区域内,利用Radon变换对图像进行旋转校正。实验结果表明,水印信息可以有效提取,并且得到的水印NC值都在0.96之上,说明该算法可以有效地抵抗旋转攻击,提高了水印安全性。Aiming at the problem that the embedded area of the watermark of the current digital watermarking algorithm is fixed and easy to be destroyed and the ability to resist rotation attacks is weak,a blind watermarking algorithm based on Radon transform and SIFT feature extraction is proposed. The SIFT algorithm is used to extract the features of the image,combined with the RANSAC algorithm to filter out stable feature points,a watermark embedding area of a suitable size is selected according to the feature points,and the watermark information is embedded in the selected area through DWT and SVD.Use Radon transform to perform rotation on the image. Experimental results show that the watermark information can be extracted effectively,and the NC value of the watermark is above 0.96. It shows that the algorithm can effectively resist the rotation attack,and the security of the watermark is improved.

关 键 词：尺度不变特征 奇异值分解 盲水印 旋转攻击 

分 类 号：TN911.7[电子电信—通信与信息系统]