| 基于SOM神经网络的城市土地覆盖遥感分类研究 |
| |
| 引用本文： | 刘艳杰,曾永年.基于SOM神经网络的城市土地覆盖遥感分类研究[J].测绘与空间地理信息,2012(6):42-45,48. |
| |
| 作者姓名： | 刘艳杰 曾永年 |
| |
| 作者单位： | 中南大学地球科学与信息物理学院 |
| |
| 基金项目： | 国家自然科学基金项目(40771198);湖南省自然科学基金项目(08 JJ6023)资助 |
| |
| 摘 要： | 土地覆盖及其变化的研究作为区域及全球环境变化研究所需的极为重要的地表参数,是遥感应用分析的主要内容之一。以往所采用的遥感分类方法主要针对侧重于土地社会属性的土地利用类型的分类研究且很难获得理想的精度。本文在非监督的自组织映射神经网络的基础上进行了一定的改进,构建了有监督的神经网络模型,以湖南省长沙市主城区的土地自然属性为侧重点,对其土地覆盖进行分类。实验结果表明:利用本文所使用的方法得到的分类结果,其总体精度和Kappa系数均高于传统的分类方法得出来的分类结果。
|
| 关 键 词： | 土地覆盖 自组织映射 神经网络 影像分类 |
Research on Remote Sensing Classification of Urban Land Coverage Based on SOM Neural Network |
| |
| LIU Yan-jie,ZENG Yong-nian.Research on Remote Sensing Classification of Urban Land Coverage Based on SOM Neural Network[J].Geomatics & Spatial Information Technology,2012(6):42-45,48. |
| |
| Authors: | LIU Yan-jie ZENG Yong-nian |
| |
| Institution: | (School of Geosciences and Info-Physics,Central South University,Changsha 410083,China) |
| |
| Abstract: | As a crucial land surface parameter needed by the research on regional and global environmental changes,the research on land coverage and the changes thereof is the main object of remote sensing application analysis.The remote sensing classification method adopted previously emphasizes on the social attribute of land utilization classification research and is hard to achieve high precision.This paper improves the unsupervised SOM neural network to set up the model of supervised SOM neural network,and classifies the land coverage in the main urban area of Changsha city which emphasizes on the natural attribute of the region studied.The result of the experiment shows that classification result achieved by using the method in this paper is higher in overall precision and Kappa coefficient than the traditional methods. |
| |
| Keywords: | land coverage SOM neural network image classification |
| 本文献已被 CNKI 等数据库收录！ |
