| 引用本文: | 宋茜,胡琼,陆苗,余强毅,杨鹏,史云,段玉林,吴文斌.农作物空间分布遥感制图发展方向探讨[J].中国农业资源与区划,2020,41(6):57~65 |
| |
|
| 摘要: |
| ［目的］农作物空间分布是农作物长势监测、病虫害监测和产量估测的基础数据，也是区域农作物种植结构优化调整、水土资源管理和政府宏观决策的重要支撑。近年来，遥感技术日益成为农作物空间分布制图的主要技术手段。［方法］文章从农作物遥感制图的现状与瓶颈出发，系统梳理了农作物空间分布遥感制图的未来发展趋势、重点方向以及存在的问题。［结果］总体看，未来农作物空间分布遥感制图呈现7个方面的发展趋势，即制图目标呈现多元化、制图单元从像元发展到地块、样本信息采集方法从线下发展到线上线下结合、众源制图数据的协同利用、制图方法从自动学习发展到深度学习； 同时，标准化的农作物空间分布数据产品的研制及共享与信息服务也是未来需要重点关注的内容。［结论］面向当前和今后相当长时间内对高精度、高时效的农作物空间分布数据的迫切需求，农作物遥感制图需要从系统性和整体性的科学视角来开展综合研究，重点加强地面数据获取、多源数据协同、智能分类算法高效运行等关键技术研究，提升农作物空间分布数据产品生产能力。 |
| 关键词: 农作物遥感制图数据方法 |
| DOI： |
| 分类号:P967 |
| 基金项目:中央级公益性科研院所基本科研业务费专项“地下水漏斗区季节性休耕地块识别方法”（1610132019027）； 国家自然科学基金项目“多云多雨区水稻熟制信息高精度提取方法”（41801371） |
|
| PROSPECT OF CROP MAPPING |
|
Song Qian1, Hu Qiong2, Lu Miao1, Yu Qiangyi1, Yang Peng1,Shi Yun1, Duan Yulin1, Wu Wenbin1※
|
|
1. Key Laboratory of Agricultural Remote Sensing, Ministry of Agriculture and Rural Affairs /Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural Sciences, Beijing 100081, China;2. College of Urban and Environmental Sciences, Central China Normal University, Wuhan, Hubei 430070, China
|
| Abstract: |
| Producing accurate crop maps is essential for numerous monitoring applications, crop area estimation, yield modeling and water demand, as well as agricultural decision support and crop pattern adjustment. From the perspective of agricultural land systems, this paper systematically proposed the progresses, the important direction and existing problems in crop mapping. It was proposed that crop mapping includes five main task, i. e., from pixel to parcel for mapping unit, from offline to online for collecting sample information, from machining learning to deep learning for classification methods, remote sensing data combination, diversity of mapping objectives and so on. The data products and data sharing are also the hot and focus of crop mapping in the future. For satisfying the urgent development needs of current and future, crop mapping should be focused on the integrated studies from a systematic and holistic view, and ground data acquisition, the integration of satellite, efficient classification algorithm will be the key development direction in the future to improve the production capacity of crop data products. |
| Key words: crop remote sensing mapping data method |
