摘要
近年来,随着经济社会发展和人类活动加剧,水资源短缺、生态环境恶化、自然灾害频发等问题对于农作物的生长具有较大影响,因此准确地提取农作物显得尤为重要。本文基于2016年Landsat-8 OLI全生育期的时序遥感影像数据,融合金塔县农作物的物候特征和地表纹理特性,提取主要农作物小麦、玉米、棉花和甜菜的时间序列NDVI曲线,探讨其时序变化特征,构建决策树分类规则,形成多维度分层次的提取方法,将其应用在2015年农作物种植结构提取中,并结合研究区2015年的Global Land Cover和统计年鉴数据对提取结果进行验证。结果表明:1)该方法可以较为准确地捕获农作物分布信息,总体精度达94.56%,Kappa系数为0.904 5,提取精度较高;2)研究区的农作物基本覆盖整个研究区域,其播种面积依次为甜菜5 540 hm^(2)、玉米4 000 hm^(2)、小麦2 270 hm^(2)、棉花300 hm^(2)。时序植被指数变化特征可以较为准确地捕获农作物信息,为精细作物分类提供了新思路,为当地决策提供农作物信息服务和基础数据支持。
In recent years, with the deepening of economic and social development and human activities, problems such as water shortage, ecological environment deterioration, and frequent natural disasters have a great impact on the growth of crops. Therefore, it is particularly important to accurately extract crops. Based on the time-series remote sensing image data of the entire growth period of Landsat8 OLI in 2016 and the crop phenological characteristics and surface texture characteristics of Jinta county, the time-series NDVI curves of main crops including wheat, corn, cotton and beet were extracted, and the time-series variation characteristics were discussed. The decision tree classification rules were constructed and the multi-dimensional and hierarchical extraction method was formed. This method was applied to the extraction of crop planting structure in 2015 and the extraction results were verified with the data of Global Land Cover and Statistical Yearbook in 2015 in the research area. Results show:(1) The method could capture crop distribution information accurately with the overall accuracy of 94.56% and Kappa coefficient of 0.904 5, indicating that the extraction accuracy was high.(2) The crops in the study area covered the whole study area. The sown area was 5.54 thousand hectares of beet, 4.00 thousand hectares of corn, 2.27 thousand hectares of wheat and 0.30 thousand hectares of cotton. The change characteristics of time-series vegetation index can capture crop information accurately, which provides a new idea for fine crop classification and crop information service and basic data support for local decision-making.
作者
石莹
穆岑
田艳君
黄月如
郭润潇
孙晓雪
SHI Ying;MU Cen;TIAN Yanjun;HUANG Yueru;GUO Runxiao;SUN Xiaoxue(College of Surveying and Mapping and Geography,Liaoning University of Technology,Fuxin 123000,China)
出处
《测绘与空间地理信息》
2022年第2期74-78,81,共6页
Geomatics & Spatial Information Technology
基金
大学生创新创业训练计划项目(201910147010,201910147057)资助。
