计算机科学 ›› 2011, Vol. 38 ›› Issue (6): 142-144.

• 计算机网络与信息安全 • 上一篇    下一篇

一种基于小波变换和神经网络的数字水印算法

陈军,张伟,杨华千,何春筱   

  1. (重庆教育学院教务处 重庆400067);(重庆教育学院院长办公室 重庆400067);(重庆教育学院网络中心重庆400067);(重庆教育管理学校计算机教研室 重庆400066)
  • 出版日期:2018-11-16 发布日期:2018-11-16
  • 基金资助:
    本文受重庆市自然科学基金项目(CSTC,2009BB2378)资助。

Novel Digital Watermark Scheme Based on Wavelet Transform and Neural Network

CHEN Jun,ZHANG Wei, YANG Hua-qiang,HE Chun-xiao   

  • Online:2018-11-16 Published:2018-11-16

摘要: 提出了一种新的基于小波变换和神经网络的数字水印算法。该算法首先把原始图像分成8X8像素的小块,根据Logistic映射产生的混沌序列选择相应的小块构造一个原始图像的子图,接着把DWT变换作用在这个子图上,得到两个子带LH:和HL1,然后把水印信息嵌入在这两个子带上,接着通过IDWT变换重构子图,最后按构成子图的顺序把每一个8X8像素的小块放回到原图中相应位置,从而得到一个嵌入了水印信息的图像。在水印的提取过程中使用了经典的I3P神经网络。实验结果表明,通过该算法嵌入的水印具有较好的鲁棒性、安全性和不可感知性。

关键词: 数字水印,神经网络,小波变换,混沌

Abstract: A novel digital watermark algorithm based on wavelet transform and neural network was proposed. Firstly,the original image was divided into some small blocks (8 X 8 pixels). The sub-image consists of some small blocks that are selected throuth logistic map. Two sub-bands LHl and HLl were generated by acting DWT on the sub-image. Secondly,the watermark was embedded into the sub-band's wavelet coefficients. The watermarked sub-image was reconstructed using CWT. Finally, we put every small block (8 X 8 pixels) of the watermarked sub-image into the initial position of the original image. Thus,we got the watermarked image. Experiments show that the watermark embedded by this scheme has good robustness,imperceptibility and security.

Key words: Watermark, Neural network, Wavelet transform, Chaos

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!