基于PSO-ORB-DWT-SVD的鲁棒水印算法  

A Robust Watermarking Algorithm Based on PSO-ORB-DWT-SVD

作　　者：刘振帅 王树梅[1] LIU Zhenshua;WANG Shumei(School of Computer Science and Technology,Jiangsu Normal University,Xuzhou 221116,China)

机构地区：[1]江苏师范大学计算机科学与技术学院,江苏徐州221116

出　　处：《湖南理工学院学报（自然科学版）》2024年第4期15-20,共6页Journal of Hunan Institute of Science and Technology(Natural Sciences)

基　　金：江苏师范大学自然科学基金(22XFRS040)。

摘　　要：针对水印算法中存在的水印不可见性与鲁棒性之间难以权衡的问题,提出一种结合离散小波变换(DWT)和奇异值分解(SVD),并融合粒子群优化(PSO)与定向快速旋转(ORB)的鲁棒水印算法.在水印嵌入阶段,首先对载体进行分块,然后对每个子块做一级DWT,选择其LL分量做二级DWT,再选择变换后的LL分量做SVD得到U矩阵,选取U矩阵第一列,利用PSO选择最佳位置进行水印的嵌入;在受到攻击时,利用ORB提取特征点进行矫正.实验结果表明,所提算法具有较好的水印不可见性,且对于噪声、滤波、压缩、旋转和剪切等多种攻击表现出较强的鲁棒性.Addressing the challenging trade-off between invisibility and robustness in watermarking algorithms,a novel robust watermarking algorithm is proposed,which combines discrete wavelet transform(DWT)and singular value decomposition(SVD),and integrates particle swarm optimization(PSO)with oriented fast and rotated brief(ORB).In the watermark embedding stage,the host image is first blocked,and then each sub-block undergoes a one-level DWT.The LL component is selected for a second-level DWT,and the transformed LL component is further processed by SVD to obtain the U matrix.The first column of the U matrix is selected,and PSO is utilized to choose the optimal location for watermark embedding.When attacked,ORB is employed to extract feature points for correction.Experimental results demonstrate that the proposed algorithm exhibits good invisibility and strong robustness against various attacks such as noise,filtering,compression,rotation,and cropping.

关 键 词：数字水印 离散小波变换 奇异值分解 粒子群优化算法 鲁棒性 

分 类 号：TP309.7[自动化与计算机技术—计算机系统结构]