【作者】 张力； 肖薇薇； 纪震； 张基宏；

【Author】 ZHANG Li,XIAO Weiwei,JI Zhen,ZHANG Jihong Faulty of Information Engineering,Shenzhen University,Shenzhen,China,518060

【机构】 深圳大学信息工程学院；

【摘要】 水印检测错误会严重损害水印系统的可信度。在水印实用化过程中,一个必须解决的问题是如何使水印检测器的错误率达到最小。本文提出在水印检测过程中采用智能信息处理技术—独立分量分析技术来实现智能水印检测器。利用独立分量分析的盲源分离能力,在水印检测过程中不需要任何关于原始图象、水印信号以及可能的攻击信息的情况下不仅可以检测到水印信号的存在而且可以精确的提取水印信号,提取的水印精确度取决于水印与原始图象、密钥之间的独立程度。实验过程中的攻击由Stirmark产生,实验数据证明这种智能水印检测器对Stirmark提供的各种攻击(如剪切、滤波、图象压缩、删行、删列、旋转、尺度变换以及多种攻击同时存在时都具有很好的鲁棒性,还可以进行多个水印的提取。这种新型的智能水印检测器的性能比现有的检水印测器优越很多。更多还原

【Abstract】 The detection errors undermine the credibility of a watermarking system.It is crucial to minimize the false detection rate in watermark detection prior to the real world applications of image watermarks.Different from early heuristic methods,the intelligent watermark decoder is designed in this paper,which adopts independent components analysis.The intelligent decoder can be extracted correctly not merely be detected without any information about the original image,watermark,attack and the accuracy of watermarking extraction depends on the key and statistical independence between the original image,watermark.Experimental results demonstrate that the proposed intelligent watermarking technique is robust with respect to attacks produced by Stirmark,such as cropping,filtering,image compression,rotation,scaling and it is even has a good robustness against combination of several attacks.This new intelligent decoder has superior advantages over the existing ones in many aspects.更多还原