摘要
高光谱遥感图像作为一种新型的遥感图像，鉴于传统的遥感图像识别方法对这种图像的识别精度较低，该文采用BP神经网络方法对高光谱遥感图像进行识别和分类，并使用赤铁矿等六种矿石的光谱图像对神经网络进行洲练，得到很好的效果。
Hyperspectral remote sensing image is a kind of new type of remote sensing images. In this paper, the BP neural network method was using in the identification and classification of hyperspectral remote sensing image for the low identification accuracy of traditional classification methods, and it receive better effect by using the six mineral samples to train the BP neural net-work.
作者
宋俊杰
葛志广
韩璞
SONG Jun-jie, GE Zhi-guang, HAN Pu （1.China University of Geosciences, Beijing 100083, China; 2.Nanyang Institute of Technology, Nanyang 473005, China）
出处
《电脑知识与技术》
2009年第11期8795-8795,8806,共2页
Computer Knowledge and Technology
关键词
高光谱图像
遥感图像
神经网络
识别
分类
hyperspectral image
remote sensing images
neural network
identification
classification
