【作者】 肖薇薇； 张力； 纪震； 张基宏；

【Author】 XIAO Weiwei,ZHANG Li,JI Zhen,ZHANG Jihong Faculty of Engineering Information,Shenzhen Univ.Shenzhen,GD 518060

【机构】 深圳大学信息工程学院；

【摘要】 本文提出以混沌吸引子作为水印信号的数字水印算法,由混沌吸引子生成的混沌水印信号具有类噪声特性,而其独特的分形自相似结构,及混沌动力学特性,使其作为水印信号比噪声类水印信号更具优良特性。文中以Lorenz混沌吸引子为例,对水印信号生成及检测算法进行了探讨。通过对混沌吸引子水印信号的主分量谱及其分形特性的分析,提出多种水印检测方案,可根据应用场合及攻击特点采用联合检测方法,以实现最优检测。实验给出混沌吸引子水印信号与高斯噪声类水印信号在同样条件下的抗攻击性能比较,实验结果表明,该算法性能上具有明显的优势。更多还原

【Abstract】 A digital watermarking algorithm based on chaotic attractor is proposed in this paper.The chaotic attractor watermark signal is noise-like,however,it has better performance than Gauss noise watermark signal because of its particular fractal self-similar structure and chaotic dynamics characteristics.Lorenz chaotic attractor watermarking generation and detection algorithms are discussed in detail as an example.Several watermarking detection methods are presented according to the principal component analysis and fractal structure of chaotic attractor.Furthermore,the united detection scheme is exploited to get best detection aiming to different application and various attacks.Experiments are developed to compare the robustness between chaotic watermarking algorithm and Gauss noise watermarking algorithm.Experimental results demonstrate that this proposed watermarking algorithm has better performance.更多还原