| 基于小波变换和RBF神经网络的交通标志识别 |
| |
| 引用本文： | 郑芳平,王伟智.基于小波变换和RBF神经网络的交通标志识别[J].福州大学学报(自然科学版),2007,35(3):397-401. |
| |
| 作者姓名： | 郑芳平 王伟智 |
| |
| 作者单位： | 福州大学自动化研究所,福建,福州,350002 |
| |
| 基金项目： | 福建省自然科学基金资助项目(A0510010)，国家自然科学基金资助项目(60675058) |
| |
| 摘 要： | 提出一种基于小波变换特征提取及采用两级神经网络分类器的交通标志识别方法.使用小波变换对图像进行处理,消除图像像素间的相关性,提取图像的整体特征作为神经网络分类器的输入向量.因交通标志类型较多,采用两级神经网络结构进行识别,图像特征先送入第一级分类器得到图像的粗分类型,再送入相应的二级子分类器进行细分.实验结果表明,这种方法具有良好的效果.
|
| 关 键 词： | 交通标志 模式识别 小波变换 径向基神经网络 两级分类器 |
| 文章编号： | 1000-2243(2007)03-0397-05 |
| 修稿时间： | 2006年12月6日 |
An approach to recognize road signs based on RBF NN using wavelet theories |
| |
| ZHENG Fang-ping,WANG Wei-zhi.An approach to recognize road signs based on RBF NN using wavelet theories[J].Journal of Fuzhou University(Natural Science Edition),2007,35(3):397-401. |
| |
| Authors: | ZHENG Fang-ping WANG Wei-zhi |
| |
| Institution: | (Institute of Automation,Fuzhou University,Fuzhou,Fujian 350002,China) |
| |
| Abstract: | Proposes an approach to recognize road signs based on a two-level RBF NN-classifier using wavelet transformation features.Wavelet transformation is employed as the image processing tool to eliminate the correlation between the pixels and extract its overall features which are used as the input vector of the NN-classifier.Because of the large number of the image types,a two-level RBF NN-classifier is developed.The image features are firstly sent to the Level 1 classifier to get the images' rough categorization,and then input to the corresponding Level 2 classifier to recognize the images' types.The experimental results show that this method works well. |
| |
| Keywords: | road signs pattern recognition wavelet transformation RBF NN two-level classifier |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《福州大学学报(自然科学版)》浏览原始摘要信息 |
| 点击此处可从《福州大学学报(自然科学版)》下载免费的PDF全文 |
