| ENVI遥感图像监督分类方法比较 |
| |
| 引用本文： | 贾建峰.ENVI遥感图像监督分类方法比较[J].西部资源,2014(6):133-136. |
| |
| 作者姓名： | 贾建峰 |
| |
| 作者单位： | 内蒙古自治区地质测绘院; |
| |
| 摘 要： | 鉴于遥感监督分类方法的普遍应用,本文介绍了五种常用的监督分类方法:平行六面体法、最小距离法、马氏距离法、最大似然法和人工神经网络分类法。就同一地区TM影像应用这五种方法进行了土地利用分类,对比分析了这五种方法的分类精度,发现人工神经网络对土地覆盖与利用的分类精度高于最大似然法,最大似然法分类精度优于平行六面体法、最小距离法和马氏距离法。所得结论对有关遥感图像分类工作具有指导和借鉴意义。
|
| 关 键 词： | 遥感 图像分类 平行六面体法 最大似然法 人工神经网络分类 |
Compare of ENVI Remote Sensing Image Classification Methods |
| |
| JIA Jian-feng.Compare of ENVI Remote Sensing Image Classification Methods[J].Western Resources,2014(6):133-136. |
| |
| Authors: | JIA Jian-feng |
| |
| Institution: | JIA Jian-feng( Geological mapping Institute in Inner Mongolia Hohhot 010020) |
| |
| Abstract: | In view of Remote Sensing Supervised Classification method used widely, this paper introduces five kinds of commonly used Supervised Classification methods： Parallelepiped, Minimum Distance, Mahalanobis Distance, Maximum Likelihood and Artificial Neural Network classification method. Using five methods to solve land use classification with TM image of same region, by comparing and analyzing the accuracy of five classification methods, we found that the accuracy of Artificial Neural Network in land cover classification is higher than that of the method of Maximum Likelihood, Maximum Likelihood classification accuracy is better than that of Parallelepiped, Minimum Distance and Mahalanobis Distance method. The conclusions may have guidance and reference significance for the Remote Sensing Image Classification work. |
| |
| Keywords: | Remote Sensing Image Classification Parallelepiped method Maximum Likelihood method Artificial Neural Network Classification |
| 本文献已被 CNKI 维普 等数据库收录！ |
