基于PCNN与粗集理论细胞图像二重数字水印算法.docVIP

基于PCNN与粗集理论细胞图像二重数字水印算法 　　摘 要:结合脉冲耦合神经网络与粗集理论,提出了一种空频域结合的二重水印算法,用于生物细胞图像的版权保护。首先在小波变换域中用粗集理论的方法找出水印嵌入位置,嵌入第一重水印信息,在此基础上,找出图像空间域中纹理复杂区域再次嵌入水印;然后讨论了水印的提取方法;最后进行了鲁棒性检测。仿真表明,该算法具有较强的不可见性和较好的鲁棒性,能准确地提取出水印信息,可以有效地对生物细胞图像进行版权保护。 　　 　　关键词:脉冲耦合神经网络; 粗集理论; 多重数字水印; 小波变换; 鲁棒性 　　中图分类号:TP399 　　文献标志码:A 　　 　　文章编号:1001-3695(2010)02-0721-05 　　doi:10.3969/j.issn.1001-3695.2010.02.088 　　 　　Double digital watermarking algorithm for cell images based onPCNN and rough set theory 　　 　　LIN Dong-mei??1,2??, MA Yi-de2, ZHANG Bei-dou2 　　 　　(1.College of Electrical Information Engineering, Lanzhou University of Technology, Lanzhou 730050, China; 2.School of InformationScience Engineering, Lanzhou University, Lanzhou 730000, China) 　　 　　Abstract:Combining pulse coupled neural network (PCNN) and rough set theory can get good achievements in digital image processing because of PCNN’s biology background and rough set theory’s excellent ability to categorize difference things. After giving the analysis of the need for biological cell images to be copyright protected, presented a new spatial-frequency-domain-based double watermarking algorithm. Firstly, used the method of rough set theory to find the location suitable for watermarking in the wavelet transform domain, and embedded the first watermarking. Then, found the regions full of textures in the spatial domain of the image with the first watermarking embedded, to embed the second watermarking. Secondly, gave a brief discussion on the watermarking extraction. Finally, also presented the robustness detection. Experimental results indicate that the algorithm has good invisibility and strong robustness. It is a feasible and effective digital image watermarking algorithm. 　　Key words:pulse coupled neural network; rough set theory; multiple digital watermarking; wavelet transform; robustness 　　 　　0 引言 　　脉冲耦合神经网络(PCNN)作为第三代人工神经网络的重要标志,在图像处理方面蕴涵着巨大潜力,在该领域做的大量研究表明,PCNN可用于图像融合、平滑、分割、识别以及边缘检测等处理[1]。目前,粗集理论已成为人工智能领域的研究热点,引起了越来越多科研人员的关