摘要
以欧空局免费资源哨兵1号(Sentinel-1B)时间序列数据为基础,利用不同农作物的后向反射系数——时间序列特征建立随机森林分类器,进行旱地作物分类,探索雷达遥感监测旱地农作物结构的可行性和实用性。研究结果表明:在三江平原地区利用多时相VV极化雷达数据可以区分玉米和大豆作物,分类精度介于65%—68%之间,利用本研究方法可以掌握农作物种植结构分布概况。
This study is based on the ESA free resources sentinel 1(Sentinel-1B)time series data,a random forest classifier was established by using the backward reflection coefficient time series characteristics of different crops to classify dry land crops and explore the feasibility of radar remote sensing monitoring dry land crop structure,The results show that corn and soybean can be distinguished by using multi-phase VV polar metric radar data in Sanjiang Plain,and the classification accuracy is between 65%—68%.The general situation of crop planting structure distribution can be mastered by using this research method.
作者
朱凤敏
吴迪
杨佳琪
ZHU Fengmin;WU Di;YANG Jiaqi(Heilongjiang Surveying and Mapping Measuring Instrument Verification Station,Harbin 150081,China;Heilongjiang Geomatics Centre,Ministry of Natural Resources,Harbin 150081,China;Heilongjiang Institute of Geographic Engineering,Harbin 150081,China)
出处
《测绘与空间地理信息》
2020年第5期105-108,共4页
Geomatics & Spatial Information Technology
作者简介
朱凤敏(1968-),女,黑龙江北安人,高级工程师,学士,主要从事测绘仪器检定、测绘生产项目技术管理方面的工作。
