| 基于神经网络计算的子波变换实现 |
| |
| 引用本文： | 曾黄麟.基于神经网络计算的子波变换实现[J].四川理工学院学报(自然科学版),1995(2). |
| |
| 作者姓名： | 曾黄麟 |
| |
| 作者单位： | 四川轻化工学院电子工程系 |
| |
| 摘 要： | 本文首先阐明了子波变换理论利用可变时域和频域分辨特性表达信号、获取信号特征的原理，提出了利用神经网络计算方法计算子波变换系数的新方法，并讨论了与人视觉系统密切相关的子波变换神经元的特性，提供了计算机仿真子波变换信号的结果。理论分析和实验结果表明，子波变换不仅具有对信号多分辨表达的特性，而且可以有效地压缩被处理信号。
|
| 关 键 词： | 神经网络 子波变换 信号近似 |
A WAVELET TRANSFORM BASED ON NEURAL NETWORK COMPUTATION |
| |
| Zeng Huanglin.A WAVELET TRANSFORM BASED ON NEURAL NETWORK COMPUTATION[J].Journal of Sichuan University of Science & Engineering:Natural Science Editton,1995(2). |
| |
| Authors: | Zeng Huanglin |
| |
| Institution: | Dept.of Electron.Eng.Sichuan Institute of Light Ind.& Chem.Tech.Zigong 643033 |
| |
| Abstract: | Clarified the rationale that wavelet transform represents a signal with varying frequency and spatial resolution and thus captures the essence of a signal,a kide of neural network computation for finding out the projection coefficients of wavelets is proposed. Different wavelet transform neurons which closely match the human vision system ate discussed.The results of several experimental simulation of wavelet transforms are presented.Theoritical analysis and experimental results show a usefulness of neural network for computing wavelet transform which,besides multi resolution reptesentation,also effectively compresses data dimension of signals. |
| |
| Keywords: | neural networks wavelet transforms signal approximation |
| 本文献已被 CNKI 等数据库收录！ |
