摘要
采用Mahalanobis Distance分类法提取了冬小麦种植面积,通过搭建决策树结构进行了不同灌溉类型冬小麦种植面积的提取,通过两个年份不同生育时期MODIS-NDVI的比较,分析了NDVI时间曲线与冬小麦长势的响应规律和水旱地冬小麦年同期长势。结果表明:2007年临汾地区冬小麦总种植面积为234778.5hm2,提取精度为96.96%,其中水地种植面积为107488.3hm2,提取精度为86.15%,旱地冬小麦为127290.2hm2,提取精度为86.16%。生育期内,水地冬小麦整体长势好于旱地冬小麦,通过NDVI变化斜率比较,表明随着生育期的推进,水地冬小麦NDVI达到峰值前的上升速度远大于旱地冬小麦,峰值后水地冬小麦NDVI下降速度小于旱地冬小麦。
In this study,Mahalanobis Distance classification was selected to extract the planting area of winter wheat. The planting area of winter wheat under different irrigation patterns was extracted through building decision tree structure.By contrasting of MODIS-NDVI at the different stages for two years,the NDVI temporal profiles and the response law of growth situation of winter wheat were analyzed.The results show that the total planting area of winter wheat was 234778.5 hm 2 with a precision of 96.96%,the pl...
出处
《农业工程学报》
EI
CAS
CSCD
北大核心
2009年第3期103-109,313,共8页
Transactions of the Chinese Society of Agricultural Engineering
基金
山西省科技攻关项目(2006031114)
山西省气象局开放式研究基金项目(SX053001)
关键词
遥感
监测
冬小麦
面积提取
长势
目视解译
决策树
remoting sensing
monitoring
winter wheat
area extraction
growth situation
visual interpretation
decision tree
