摘要
利用卫星遥感数据制作复杂地形环境的植被图面临的最主要问题是精度，单纯对遥感数据（ＴＭ或ＳＰＯＩ）进行监督或非监督分类的精度低于５０％。本文选择美国亚利桑那州ＳａｎｔａＣａｔａｌｉｎａ山脉的ＰｕｓｃｈＲｉｄｇｅ作为研究区，分析地理信息系统模型在改善植被分类精度中的作用。结果表明，通过结合辅助数据和应用地理信息系统模型，其精度可以从３７．４１％提高到７１．６７％（ＳＰＯＴ数据，非监督分类），或从５０．０７％提高到６１．５０％（ＴＭ数据，监督分类）。同时表明用ＳＰＯＴ数据进行山区植被制图的效果好于ＴＭ数据。
The classification accuracy of vegetation mapping by satellite ima- gery in a complex terrain environment can be improved by using ancillary data andimagery spatial features extracted from the images(Richord,1992).Thls study isto test the role of GIS spatial and spectral analysis model in aiding the classifica- tion of satellite data. Three GIS programs is developed,which improve the accu- racy of unsupervised classification for SPOT data from 37.41% to 71.67%. Second objective is to test the ability of two satellite system,SPOT and LandsatThematic Mapper(TM),in mapping of vegetation in mountain region,Both data areprocessed with supervised classification in-corporating with ancillary data. The ac- curacy with SPOT data is higher than with TM.
出处
《环境遥感》
CSCD
1995年第1期30-37,共8页
基金
世界实验室资助培训计划
瑞士世界实验室ICSC资助黄河下游实时洪水监测和洪水管理系统项目
关键词
遥感
GIS
植被
制图
环境遥感
Vegetation Mapping,GIS Model,Ancillarv data
