河南农业科学 ›› 2022, Vol. 51 ›› Issue (1): 154-161.DOI: 10.15933/j.cnki.1004-3268.2022.01.019

• 农业信息与工程·农产品加工 • 上一篇    下一篇

基于无人机遥感影像的多品种水稻抽穗时间估算

莫佳才,彭漪,刘小娟,王靖   

  1. (武汉大学遥感信息工程学院,湖北 武汉 430079)
  • 收稿日期:2021-05-24 出版日期:2022-01-15 发布日期:2022-03-18
  • 通讯作者: 彭漪(1984-),女,湖北武汉人,副教授,博士,主要从事定量遥感研究。E-mail:ypeng@whu.edu.cn
  • 作者简介:莫佳才(1996-),男,广西玉林人,在读硕士研究生,研究方向:定量遥感。E-mail:wdmjcc@whu.edu.cn
  • 基金资助:
    国家自然科学基金面上项目(41771381)

Estimation of Heading Time for Rice of Different Cultivars Based on UAV Remote Sensing Images

MO Jiacai,PENG Yi,LIU Xiaojuan,WANG Jing   

  1. (School of Remote Sensing and Information Engineering,Wuhan University,Wuhan 430079,China)
  • Received:2021-05-24 Published:2022-01-15 Online:2022-03-18

摘要: 为了实现用遥感方法监测并估算多品种水稻的抽穗时间,利用不同气候区域1 102个水稻品种的多时相无人机多光谱遥感影像数据,分析水稻光谱与植被指数,使用单高斯函数对原始多时相归一化差分红边植被指数(Normalized difference red edge index,NDRE)数据进行拟合,提取拟合曲线特征峰值。结果表明,峰值时间与实际抽穗时间的Peason相关系数达0.906,由此建立多品种水稻抽穗时间遥感估算模型IHDDAS=1.3×PEAKDAS-18(其中,DAS为播种后天数,IHDDAS 为始穗日期的DAS,PEAKDAS 为曲线峰值DAS)。模型实现了实际抽穗时间的准确估计,在建模区1 014个品种的均方根误差(RMSE)为3.15 d,在品种数为48、40个的2个验证区的RMSE分别为5.02、2.99 d。该模型对不同地点、气候、品种均具有较好的适用性。

关键词: 水稻, 抽穗时间, 无人机遥感, 植被指数

Abstract: For the purpose of remotely monitoring and estimating heading time for rice of different cultivars,based on the multi‐temporal and multi‐spectral UAV remote sensing images of 1 102 rice cultivars under different climatic regions,we analyzed rice spectrum and vegetation index,fit the original multi‐temporal normalized difference red edge index(NDRE)data using single Gaussian function,and extracted the characteristic peak value of fitting curve. The results showed that the Pearson correlation coefficient between peak time and heading time reached 0. 906,based on which we established the remote sensing estimation model of multi‐cultivar rice heading time IHDDAS=1.3×PEAKDAS-18(DAS means the days after sowing,IHD DAS means DAS of initial heading date,PEAKDAS means DAS of curve peak).The model worked well for estimating the measured heading time.The root mean square error of 1 014 varieties in the modeling experiment was 3.15 d,and 5.02 d and 2. 99 d in the two verification experiments with 48 and 40 varieties,respectively. The model has good applicability to rice under different regions,climate and cultivars.

Key words: Rice, Heading time, UAV remote sensing, Vegetation index

中图分类号: