摘要
笔者以2007年11月中巴资源卫星多光谱遥感影像数据为例,选择山东省济宁市市中区作为研究区,利用ERDAS Imagiue9.2遥感影像处理软件对数据进行预处理后,运用最大似然法(Maximum Likelihood)、马氏距离法(Mahalanobis Distance)、最小距离法(Minimum Distance)进行监督分类试验.然后对分类后的结果进行主观和客观的分析,并评价各种分类结果的精度,为用户进行影像分类前选择分类方法提供参考.
Taken the CBERS multi -spectral remote sensing image data of Jining as an example, after pretreatmented with the software of ERDAS Imagine9.2, different supervised classifications ( Maximum Likelihood , Mahalanobis Distance, Minimum Distance)have been applied in image classification and accuracy assessment has been applied to it. According to the experiment result mentioned above, the paper provides a reference of image classifications to user.
出处
《山东师范大学学报(自然科学版)》
CAS
2011年第3期80-83,共4页
Journal of Shandong Normal University(Natural Science)
关键词
监督分类
分类精度
遥感图像
supervised classification
the classification accuracy
remote sensing image