摘要
从MODIS影像中划出一块无云区域作为研究对象,求出该研究区的NDVI和EVI,运用K-均值分类法分别对2种植被指数进行非监督分类.通过对2种植被指数分类结果进行比较得出:2种植被指数的分类结果在总体上是一致的,但是在局部区域,NDVI对于水体的分辨能力优于EVI,而EVI对植被则有更强的分辨能力.2种植被指数分类结果精度评价总精度为70.00%,Kappa系数为0.585.
A cloudless region extracted from a MODIS image was used as the object of this study. ＂K-mean unsupervised method＂ was used to classify the two vegetation indices, NDVI and EVI, of the studied area. A comparison of the vegetation indices thus obtained showed that the two indices were, on the whole, consistent with each other; however, NDVI gave a better resolution for water bodies and wet things while EVI had a greater capacity of identifying vegetations. They had a precision appraisal score of 70.00% and a Kappa coefficient of 0. 585.
出处
《西南师范大学学报（自然科学版）》
CAS
CSCD
北大核心
2007年第2期70-75,共6页
Journal of Southwest China Normal University(Natural Science Edition)
基金
重庆市教委科技资助项目(1-82)
