Robust Image Watermarking Technology Based on Beetle Antennae Search and LP-SWT-SVD
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
万方数据
为了有效保护数字媒体的知识产权,提出一种基于拉普拉斯金字塔和平稳小波变换(Stationary Wavelet Transform,SWT)及奇异值分解(Singular Value Decomposition,SVD)的图像水印嵌入算法.该算法首先对原始图像进行拉普拉斯金字塔分解,然后对得到的残差图像进行一级平稳小波变换,得到低频子带LL1和高频子带HH1,分别对其进行SVD分解,并将SVD分解后的水印分别嵌入低频和高频子带的奇异值矩阵中,使用天牛须算法(Beetle Antennae Search,BAS)优化水印嵌入过程.水印检测时,将从LL1和HH1子带中提取的水印进行比较,选择效果较好的作为最终结果.仿真实验与其他文献的对比分析证明该算法不可见性和鲁棒性都较好.
In order to effectively protect the intellectual property rights of digital media,an image watermark embedding algorithm based on Laplacian Pyramid(LP),Stationary Wavelet Transform(SWT),and Singular Value Decomposition(SVD)was proposed in this study.First of all,the original image was decomposed by Laplacian pyramid.After that,the residual image was decomposed by one-level SWT,the low frequency subband LL1 and high frequency subband HH1 were obtained.They were decomposed by SVD,and the watermarking after SVD was embedded in the singular value matrix of the low frequency subband and high frequency subband.The watermark embedding process was optimizing by the Beetle Antennae Search(BAS)algorithm.In watermarking detection,the watermarks extracted from LL1 and HH1 subbands were compared,and the better result was chosen as the final result.Simulation experiments and comparative analysis with other literature have shown that the algorithm has good invisibility and robustness.
Laplacian pyramidStationary wavelet transformSingular value decompositionBeetle antennae search algorithm
NETL
NSTL
万方数据
