基于融合NDVI和EVI时间序列的遥感影像分类研究  

Research on Remote Sensing Image Classification Based on NDVI and EVI Time Series

作　　者：朱满[1,2] 胡光宇[3] 于之峰[3] 

机构地区：[1]西安勘察测绘院,西安7100482 [2]武汉大学遥感信息工程学院,武汉430079 [3]武汉大学测绘遥感信息工程国家重点实验室,武汉430079

出　　处：《遥感信息》2009年第5期44-46,66,共4页Remote Sensing Information

摘　　要：通过比较分析发现同一地物的NDVI和EVI时间序列变化规律具有一定差异,据此,提出结合NDVI和EVI植被指数时间序列的分类方法,结果表明比单独使用NDVI或EVI时间序列总体精度要提高2.7%,kappa系数提高0.04。Through comparing and analyzing the time series of EVI and NDVI of the same features,we find that they have some differences in the law of change.So,a classification method combined with proposed EVI and vegetation index NDVI time series was proposed,and the results show that the method improves the overall accuracy 2.7%,and kappa coefficient 0.04 than the method which uses NDVI or EVI time series separately.

关 键 词：NDVI EVI 时间序列 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]