| 基于改进的BP神经网络模型的遥感图像分类方法研究 |
| |
| 引用本文： | 刘钦龙,焦斌亮,刘立.基于改进的BP神经网络模型的遥感图像分类方法研究[J].电光与控制,2009,16(8). |
| |
| 作者姓名： | 刘钦龙 焦斌亮 刘立 |
| |
| 作者单位： | 燕山大学信息科学与工程学院,河北,秦皇岛,066004 |
| |
| 摘 要： | 遥感图像通过像元灰度值的高低差异和空间变化来表示不同属性的地物目标及其分布情况。随着人工神经网络理论的发展,神经网络技术日益成为遥感图像分类处理的有效手段,并有逐渐取代传统遥感图像分类方法的趋势。重点分析了BP神经网络模型的遥感图像分类方法,并针对BP算法的学习速率慢做了改进,利用M atlab 7.0作为实验平台,对北京市海淀区遥感影像进行分类实验。实验结果表明,与基本BP算法相比较,改进的BP算法网络训练收敛速度快,分类方法是有效的。
|
| 关 键 词： | 遥感图像 灰度值 BP神经网络 Matlab |
On Remote Sensing Image Classification Method Based on Improved BP Neural Network Model |
| |
| LIU Qinlong,JIAO Binliang,LIU Li.On Remote Sensing Image Classification Method Based on Improved BP Neural Network Model[J].Electronics Optics & Control,2009,16(8). |
| |
| Authors: | LIU Qinlong JIAO Binliang LIU Li |
| |
| Affiliation: | Information Science and Engineering Institute;Yanshan University;Qinhuangdao 066004;China |
| |
| Abstract: | Remote sensing images indicate the physical objects with different attributes and their distribution through pixel gray value differences and the level of spatial variation.With the development of Artificial Neural Network(ANN) theory,remote sensing image classification of ANN theory became an effective means and gradually replace traditional methods of remote sensing image classification.The remote sensing image classification methods based on Back-Propagation(BP) network model was analyzed in detail,and a... |
| |
| Keywords: | remote sensing image gray value BP neural network Matlab |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
