基于RBF神经网络的空域图像水印改进算法

Reaearch on Digital Watermark Based on RBF Neural Network

文章提出了一种利用径向基函数(RBF)神经网络实现的数字图像水印的改进新算法,从理论上对新算法进行了分析,通过仿真实验进行了验证。仿真结果表明,新算法与传统基于RBF神经网络图像水印算法相比,在保证速度不减慢的基础上其抗攻击能力有了显著的改善。该算法的创新之处在于将多级神经网络的思想引入数字图像水印算法中,利用几个性能不同的网络实现一种系统的功能。

This paper presents a new image digital watermark algorithm based on RBF neural network.After much simulation research,we find that new algorithm is good at resisting attack when calculating speed is still quick.The experimental results demonstrate that watermarked image has good visual effect,which demonstrates that this algorithm is feasible.