摘要
小波神经网络有机地融合了小波分析的时频特性和神经网络自适应优点。本文将小波神经网络应用于图像表述,提出相应的图像表述算法。分别采用两种小波函数作为网络激励函数,以验证图像表述效果。实验结果表明,小波神经网络能够有效地表述图像,其算法具有较强的鲁棒性。
Wavelet neural network organically combines the good localization characteristics of the wavelet transform theory and the adaptive learning virtue of neural networks.In this paper,a wavelet neural network is applied to image representation and the correlated algorithms are also proposed.Two kinds of wavelet functions are respectively selected as the excitation function of wavelet neural network to testify the performance of image representation.The experiment results show that the wavelet neural network can effectively represent the image and the proposed algorithms have good robustness.
出处
《影像技术》
CAS
2006年第1期31-33,共3页
Image Technology
关键词
小波神经网络
图像表述
小波函数
wavelet neural network
image representation
wavelet function
