基于混沌和HVS的小波域自适应图像水印算法  

Adaptive image watermarking algorithm in wavelet domain based on chaos and HVS

作　　者：李峰[1] 陈光喜[1] 丁勇[1] 潘黄琼[1] 刘燕雄[1] 

机构地区：[1]桂林电子科技大学数学与计算科学学院,广西桂林541004

出　　处：《计算机应用研究》2012年第6期2224-2227,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(60963024);广西学位与研究生教育改革和发展专项课题资助(200910595R06)

摘　　要：针对常规的小波域水印算法自适应性差的缺陷,提出了一种新的基于混沌和人类视觉系统的小波域自适应图像水印算法。首先对载体图像进行二级DWT变换;然后对特定的中频子带进行分块,并对每块做SVD分解;最后将Logistic混沌置乱后的二值水印信息嵌入到每块的最大奇异值中,嵌入的强度根据HVS原理自适应地进行调整。仿真实验表明,该算法不仅具有很好的不可见性,且对JPEG压缩、加噪、滤波等常见的图像攻击具有很强的鲁棒性。Aiming at the defect of self-adaptability in normal digital watermarking algorithm in DWT domain,this paper proposed an adaptive image watermarking algorithm based on chaos and HVS in DWT domain.Firstly,applied a two-level DWT to the cover image.Secondly,divided a specific middle frequency sub-band into many blocks which decomposed with SVD.Finally,adaptively embedded the binary watermark image with Logistic chaos scramble into the largest singular value of each block.It self-adaptively adjusted the embedding intensity according to the quantization step of each block.Simulation results prove this algorithm has a good transparence and robustness against attacks such as JPEG compression、adding noise、filtering and other image manipulations.

关 键 词：数字水印 混沌 人类视觉系统 离散小波变换 自适应 奇异值分解 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]