| 基于四层BP神经网络模型的数字水印技术 |
| |
| 引用本文： | 王忠,肖念.基于四层BP神经网络模型的数字水印技术[J].教育技术导刊,2008(2). |
| |
| 作者姓名： | 王忠 肖念 |
| |
| 作者单位： | 武汉工程大学计算机科学与工程学院 湖北武汉430073 |
| |
| 摘 要： | 对神经网络应用于数字水印技术进行了改进,选择了一种DWT变换嵌入算法的BP四层模型作为神经网络的训练模型,它能更好地实现了水印的盲提取。实验结果表明,该算法对常用的图像处理如压缩、低通滤波、加噪、剪切、旋转和缩放等攻击具有较好的鲁棒性和不可见性。
|
| 关 键 词： | 神经网络 盲水印 鲁棒性 不可见性 |
Digital Watermarking Algorithm Based on BP Four-layer Neural Network |
| |
| WANG Zhong,XIAO Nian.Digital Watermarking Algorithm Based on BP Four-layer Neural Network[J].Introduction of Educational Technology,2008(2). |
| |
| Authors: | WANG Zhong XIAO Nian |
| |
| Abstract: | The authors improved the application method of four-layer neural network to digital watermarking technology and chose BP for neural network of DWT transformation as the training mode,which could better realize blind withdraw of watermark.The results show that this algorithm provides high robustness and invisibleness against image processing attacks such as compression,low-pass filtering,adding noise,cropping,rotating and zooming. |
| |
| Keywords: | neural network blind watermark robust invisibleness |
| 本文献已被 CNKI 等数据库收录！ |
|
