基于CNN和NSCT的零水印算法  

Zero Digital Watermarking Algorithm Based on CNN and NSCT

作　　者：赵杰[1] 屈正庚[2] 

机构地区：[1]商洛学院物理与电子信息工程系,商洛726000 [2]商洛学院计算机科学系,商洛726000

出　　处：《科学技术与工程》2013年第5期1368-1372,共5页Science Technology and Engineering

基　　金：商洛学院科研基金项目(09SKY032.09SKY033);陕西省教育厅科研计划项目(11JK1067)资助

摘　　要：提出一种基于细胞神经网络(CNN)和非抽样Contourlet变换(NSCT)的零水印算法。首先对原始载体图像进行非抽样Contourlet变换,获得其图像的低频逼近子带;然后对水印信息进行置乱,将其与图像的低频逼近子带一起输入CNN网络,得到注册图像。水印检测时可以利用尺度不变特征变换(SIFT)进行几何校正。实验结果表明,该方法可以获得较好的检测精度;同时对于加噪、滤波、JPEG压缩、剪切攻击也具有很好的鲁棒性。由于细胞神经网络对图像处理的并行性与可由硬件实现的特点,该算法可应用于实时性要求较高的场合。Watermark embedding introduces inevitably some perceptible quality degradation of the host image. Another problem is the inherent conflict between imperceptibility and robustness. However, the zero-watermarking technique can extract some essential characteristics from the host image and use them for watermark registration and detection. The original image was decomposed into series of multiscale and directional subimages after nonsubsampled contourlet transform （NSCT）. The low-frequency subimage and watermark image are inputs of the cellular neural network（CNN）, and the zero-watermarking registration image is the output. The scale-invariant feature transform（SIFT） is employed to against scaling and rotation attacks. To investigate and improve the security and robustness, the original watermark and registration image are scrambled or encrypted. The proposed method is simple for hardware realization. Experimental results show that it is robust to many common image operations.

关 键 词：细胞神经网络 非抽样CONTOURLET变换 零水印 尺度不变特征变换 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]