安徽农业科学2025，Vol.53Issue(9)：196-202,7.DOI:10.3969/j.issn.0517-6611.2025.09.042
基于多光谱卫星影像的小麦种植分布遥感监测方法研究
Study on Remote Sensing Method of Wheat Planting Area Monitoring Based on Multispectral Satellites Images
摘要
Abstract
This study adopted a remote sensing method of wheat planting area monitoring.Based on multispectral remote sensing images,an experiment was carried out in a 440 km2 area which was part of Suzhou City.NDVI time series of Sentinel-2 satellites was established,and S-G method was used to smooth the NDVI time series.U-Net method was employed to identify wheat planting area while other two methods,SAM and MLP,were used for comparison.The result showed that all of three methods could achieve the accuracy level of 80%.U-Net had the lowest omission rate,and SAM had the lowest error rate.U-Net which achieved the overall accuracy of 83.15%,is the best method under the same condition.关键词
多光谱/归一化植被指数/U-Net模型/光谱匹配/人工神经网络Key words
Multispectral/NDVI/U-Net/SAM/MLP分类
农业科技引用本文复制引用
陈春晖,梁树能,李金超..基于多光谱卫星影像的小麦种植分布遥感监测方法研究[J].安徽农业科学,2025,53(9):196-202,7.基金项目
安徽省自然资源科技项目(2023-K-2). （2023-K-2）
