中国农业科学院农业信息研究所 中国农学会
Agricultural Information Institute of CAAS China Association of Agricultural Science Societies
En
Ch
中国农业科学院农业信息研究所 中国农学会 主办

AI检索

订阅

返回顶部
当前位置: 首页 > 文章 > 基于多时相GF-1遥感影像的作物分类提取 河南农业科学 2016,45 (1) 152-159
Position: Home > Articles > Crop Classification Extraction Based on Multi-temporal GF-1 Remote Sensing Image Journal of Henan Agricultural Sciences 2016,45 (1) 152-159

基于多时相GF-1遥感影像的作物分类提取

作  者：
贺鹏;徐新刚;张宝雷;李振海;金秀良;张秋阳;张勇峰
单  位：
国家农业信息化工程技术研究中心;山东师范大学
关键词：
遥感;作物;分区;多时相;决策树;GF-1
摘  要：
为了提高遥感影像数据对作物分类提取的精度,更多地反映作物的空间分布结构和物候差异,以黑龙江农垦赵光农场为研究对象,提出一种基于分区与决策树分层分类相结合的作物遥感分类方法,利用2014年高分一号卫星(GF-1)WFV遥感影像数据(4景)开展主要作物的识别分类提取。首先,结合实地调查与影像光谱特征信息的总体分布,将研究区分割成3个子区域(西南区、北部区和东南区);其次,基于多时相遥感影像序列,分析主要作物的反射光谱和植被指数的时序变化特征,构建基于决策树分层分类的主要作物遥感分类模型,成功提取了赵光农场主要作物的空间种植信息。结果表明,2种分类方法的精度都很高,总体精度均在97.00%以上,Kappa系数均在0.900 0以上。分区分类更优于整幅图像非分区分类,总体精度达到98.10%,Kappa系数达到0.960 7;非分区分类总体精度为97.50%,Kappa系数为0.948 3。研究表明,基于分区与决策树分类法相结合的作物分类结果精度,明显优于不使用分区分类的结果。由分区与决策树分层相结合的分类方法能够有效提高黑龙江垦区主要种植作物分类的准确性和精度。
译  名：
Crop Classification Extraction Based on Multi-temporal GF-1 Remote Sensing Image
作  者：
HE Peng;XU Xingang;ZHANG Baolei;LI Zhenhai;JIN Xiuliang;ZHANG Qiuyang;ZHANG Yongfeng;College of Population,Resource and Environment,Shandong Normal University;National Engineering Research Center for Information Technology in Agriculture;Key Laboratory of Agri-informatics,Ministry of Agriculture;Beijing Engineering Research Center of Agricultural Internet of Things;
关键词：
remote sensing;;crops;;partition;;multi-temporal;;decision tree;;GF-1
摘  要：
With Zhaoguang farm in Nongken region of Heilongjiang province as study object,a remote sensing classification method combined with area segmentation and layered decision tree was came up to improve the classification accuracy for crop using remote sensing data and reflect more spatial distribution structure of crop and phenology differences. Four GF-1 remote sensing images were used to extract the major crops. First,the study area was separated into three regions( southwestern region,north region and southeastern region) depended on the total distribution of characteristic spectrum information acquired by field survey and image; Next,time-series change features of reflection and vegetation index of main crops were analysed based on multi-temporal remote sensing image series,then a main crop classification model base on decision tree was built,and the spatial planting information of major crops in the Zhaoguang farmwas extracted. The results indicated that both of two classification methods had high precision with total accuracy higher than 97. 00% and Kappa coefficient larger than 0. 900 0. The classification result using the area-separated method( with total accuracy of 98. 10% and Kappa coefficient of 0. 960 7) was better than that not not using the area separation method( with total accuracy of 97. 50% and Kappa coefficient of 0. 948 3). The study indicated that the classification method combined with area-separation and decision tree can efficiently improved the classification accuracy and precision of main planting crops in region of Heilongjiang Nongken region.

相似文章

导出引用
计量
文章访问数: 16
HTML全文浏览量: 0
PDF下载量: 1

所属期刊

推荐期刊

友情链接Friendly link
农业信息研究所Agricultural information institute 农业智能知识服务平台Agricultural intelligent knowledge service platform 农业学术服务平台Agricultural Knowledge Service System 中国知网CNKI 维普网VIP 万方WANFANG DATA
联系我们Contact us

电话Tell：010-82109652-808

邮箱E-mail：agrijournal@caas.cn

服务Service
关于我们About us 意见建议Opinions and suggest
版权说明Copyright
单位：中国农业科学院农业信息研究所 中国农学会 © 2017中国农业科学院农业信息研究所 ICP备案号：京ICP备09089781号-46