河南农业科学 ›› 2023, Vol. 52 ›› Issue (6): 160-171.DOI: 10.15933/j.cnki.1004-3268.2023.06.017

• 农业信息与工程·农产品加工 • 上一篇    下一篇

融合主被动遥感影像的冬小麦种植面积提取研究

张科谦1,程钢1,吴微2,宋向阳1,张子谦3,姚顺1,吴帅1   

  1. (1.河南理工大学测绘与国土信息工程学院,河南 焦作 454003;2.桂林理工大学测绘地理信息学院,广西 桂林 541000;3.河南科技大学软件学院,河南 洛阳 471003)
  • 收稿日期:2023-01-08 出版日期:2023-06-15 发布日期:2023-07-11
  • 通讯作者: 程钢(1981-),男,山东莱阳人,教授,博士,主要从事地理信息理论与方法等研究。E-mail:chenggang1218@163.com
  • 作者简介:张科谦(1998-),男,河南濮阳人,在读硕士研究生,研究方向:农业遥感。E-mail:keqianzhang@home.hpu.edu.cn
  • 基金资助:
    国家自然科学基金资助项目(41001226);中国博士后科学基金项目(2015M582831);河南省高校基本科研业务费专项资金资助项目(NSFRF180329)

Extraction of Winter Wheat Planting Area Based on Fused Active and Passive Remote Sensing Images

ZHANG Keqian1,CHENG Gang1,WU Wei2,SONG Xiangyang1,ZHANG Ziqian3,YAO Shun1,WU Shuai1   

  1. (1.College of Surveying and Land Information Engineering,Henan Polytechnic University,Jiaozuo 454003,China;2.College of Geomatics and Geoinfornation,Guilin University of Technology,Guilin 541000,China;3.School of Software,Henan University of Science and Technology,Luoyang 471003,China)
  • Received:2023-01-08 Published:2023-06-15 Online:2023-07-11

摘要: 为了快速、准确地获取作物分布信息,探索使用主动遥感影像(Sentinel-1A)和被动遥感影像(Sentinel-2)提取冬小麦空间分布的可行性。首先,根据冬小麦的物候特征,合成冬小麦全生育期的Sentinel-1A影像;并依据各类地物的NDVI(归一化植被指数)时序曲线合成一期高质量的冬小麦越冬后Sentinel-2影像。其次,设计Sentinel-1A影像、Sentinl-2影像和融合Sentinel-1A与Sentinl-2主被动遥感影像3种分类方案,然后在Google Earth Engine(GEE)云平台上基于随机森林算法对冬小麦进行分类。结果表明,基于全生育期Sentinel-1A 影像的冬小麦用户精度和生产者精度分别为83.15% 和86.44%,提取结果中存在较多的“椒盐”噪声;基于冬小麦越冬后Sentinl-2影像的冬小麦用户精度和生产者精度分别为87.98%和84.75%,提取精度较使用全生育期Sentinel-1A影像有所提高,但分类结果受“异物同谱”的影响,产生许多错分;融合主被动遥感影像的冬小麦用户精度和生产者精度分别为96.57%和95.48%,相较于仅使用单一数据源,冬小麦分类精度有不同程度的提升。

关键词: 冬小麦, 种植面积, GEE, Sentinel影像, 生育期, 随机森林, 主动遥感, 被动遥感

Abstract: To obtain fast and accurate crop distribution information,the feasibility of using active remote sensing imagery(Sentinel⁃1A)and passive remote sensing imagery(Sentinel⁃2)to extract the spatial distribution of winter wheat was analysed. Firstly,Sentinel⁃1A images of winter wheat at the whole growth stage were synthesized based on the phenological characteristics of winter wheat,and high quality Sentinel⁃2 images of winter wheat after overwintering were synthesized based on the normalized vegetation index(NDVI)time series curves of various types of features.Three classification schemes,Sentinel⁃1A images,Sentinl⁃2 images and fused Sentinel⁃1A and Sentinl⁃2 active⁃passive remote sensing images,were designed,and then winter wheat was classified based on the random forest algorithm on the Google Earth Engine(GEE)cloud platform. The results showed that,the user accuracy and producer accuracy of winter wheat based on Sentinel⁃1A images at the whole growth stage were 83.15% and 86.44% respectively,and there was more“pepper”noise in the extraction results;the user accuracy and producer accuracy of winter wheat based on Sentinl⁃2 images after overwintering were 87.8% and 84.75% respectively,and the extraction accuracy was improved compared with that of Sentinel⁃1A images at the whole growth stage,but the classification results were influenced by the“same spectrum of foreign matter”,resulting in many misclassifications;the user accuracy and producer accuracy of winter wheat with fused active and passive remote sensing images were 96.57% and 95.48%,respectively,compared with that of using only a single data source,the classification accuracy of winter wheat was improved to different degrees.

Key words: Winter wheat, Planting area, GEE, Sentinel image, Growth stage, Random forest, Active remote sensing, Passive remote sensing

中图分类号: