引用本文: | 李强,王正志.基于MRF模型和人工神经网络的遥感图像分类综合方法研究[J].国防科技大学学报,1999,21(1):62-66.[点击复制] |
Li Qiang,Wang Zhengzhi.Research on NN-Based Remote Sensing Image Classification and Smoothing Integrated Technique[J].Journal of National University of Defense Technology,1999,21(1):62-66[点击复制] |
|
|
|
本文已被:浏览 6019次 下载 5259次 |
基于MRF模型和人工神经网络的遥感图像分类综合方法研究 |
李强, 王正志 |
(国防科技大学 自动控制系 湖南 长沙 410073)
|
摘要: |
本文总结了提高计算机遥感信息分类精度的四个有效途径, 据此提出了基于三维Hopfield人工神经网络模型的遥感信息分类及平滑处理综合技术。实验表明, 该方法可明显提高森林类型划分、土地利用调查等遥感应用专题的分类精度。 |
关键词: 遥感信息分类, 分类后处理, 人工神经网络 |
DOI: |
投稿日期:1998-09-09 |
基金项目:国家卫星应用重点项目资助 |
|
Research on NN-Based Remote Sensing Image Classification and Smoothing Integrated Technique |
Li Qiang, Wang Zhengzhi |
(Depart. of Automatic control, NUDT, Changsha, 410073)
|
Abstract: |
Summing up four ways to enhance the remote sensing computer classification precision, we put forward a remote sensing classification and smoothing integrated technique based on 3-D Hopfield network theory. The experiment results show that this method can improve the precision of classification saliently. |
Keywords: Remote Sensing, Classification, Post-classification, Neural network |
|
|
|
|
|