摘要
吉林省四平市梨树县作为全国粮食生产先进县、国家现代农业示范区[1],稳种植面积是该区域农业发展的首要任务。该文基于多时相Landsat8数据,结合全球高精度土地覆盖数据集GlobeLand30数据,从影像采样像元统计到典型地物光谱曲线,逐步识别研究区内的玉米种植区,从而提取吉林省梨树县玉米种植面积数据,并对比吉林省统计年鉴数据进行误差分析,最终获取准确的玉米面积提取方法。研究结果表明,使用吉林省玉米生长期中抽雄期的多时相Landsat8数据进行种植面积的识别及提取准确度较高,其次为乳熟期;多时相Landsat8数据在玉米面积的提取方面,乃至农业遥感应用方面都有一定的优势。
Lishu County in Siping City of Jilin Province,as a national advanced county in grain production and a national modern agricultural demonstration area,stabling planting area is the top priority of the region's agricultural development.This paper is based on multi-temporal Landsat8 data,combined with the global high-precision land cover dataset GlobeLand30 data,from image sampling pixel statistics to typical feature spectral profiles,the corn growing areas in the study area are identified step by step so as to extract data on corn planting area in Lishu County in Jilin Province,and compared with data from the Jilin Provincial Statistical Yearbook for error analysis and ultimately to obtain an accurate method of extracting corn planting area.The results of the study showed that the identification and extraction of planted area was more accurate using multi-temporal Landsat8 data from the mid-male staminate stage of the corn growing season in Jilin Province,followed by the lactation stage.The multi-temporal Landsat8 data has advantages in the extraction of corn area,and indeed in agricultural remote sensing applications.
出处
《科技创新与应用》
2023年第13期47-50,共4页
Technology Innovation and Application
基金
吉林省大学生创新创业训练计划项目(S202210191038)。