摘要
结合广西水稻面积的遥感解译工作,应用SPOT4遥感数据和遥感处理软件ENVI,利用面向对象的遥感分类的方法提取早稻种植面积。分类结果表明,利用面向对象的分类方法有效解决逐像素分类结果的"椒盐"效应,获得比传统的像素级分类方法更高的分类精度,为广西水稻种植面积的自动提取提供了广阔的前景。
The paper combines the remote sensing interpretation of rice area in Guangxi,applies SPOT4 remote sensing data and remote sensing software ENVI,makes use of object-oriented remote sensing classification method to extract rice planting area.The classification result shows that by using object-oriented classification method,it is helpful to solve the ＂salt and pepper＂ effect that is the result of pixel-by-pixel classification effectively,and receive higher classification accuracy compared to the traditional pixel-level classification method,and the provide the broad prospective for the automatic extraction of rice planting area in Guangxi.
出处
《遥感信息》
CSCD
2010年第1期53-56,共4页
Remote Sensing Information
基金
国家自然科学基金项目(40761027)
关键词
面向对象分类
K近邻法
水稻
广西
object-oriented classification K nearest neighbor rice guangxi
