无人机多光谱影像在稻纵卷叶螟危害监测中的应用  

Application of multispectral image taken by unmanned aerial vehicle in monitoring Cnaphalocrocis medinalis Güenée damage on rice growth

作　　者：郭铭淇 包云轩[1,2,3] 黄璐[1,2,3] 陈晨 杨荣明[4] 朱凤 GUO Ming-qi;BAO Yun-xuan;HUANG Lu;CHEN Chen;YANG Rong-ming;ZHU Feng(Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters,Nanjing University of Information Science and Technology,Nanjing 210044,China;Jiangsu Key Laboratory of Agricultural Meteorology/School of Applied Meteorology,Nanjing University of Information Science and Technology,Nanjing 210044,China;Key Laboratory of Meteorological Disaster of Ministry of Education/Joint International Research Laboratory of Climate and Environment Change,Nanjing University of Information Science and Technology,Nanjing 210044;Plant Protection Station in Jiangsu Province,Nanjing 210013,China)

机构地区：[1]南京信息工程大学气象灾害预报和评估协同创新中心,江苏南京210044 [2]江苏省农业气象重点实验室/南京信息工程大学应用气象学院,江苏南京210044 [3]南京信息工程大学气象灾害教育部重点实验室/气候与环境变化国际合作联合实验室,江苏南京210044 [4]江苏省植物保护站,江苏南京210013

出　　处：《江苏农业学报》2023年第7期1530-1542,共13页Jiangsu Journal of Agricultural Sciences

基　　金：国家自然科学基金面上项目(41975144);江苏省重点研发项目(BE2019387)。

摘　　要：为探讨利用无人机多光谱影像监测稻纵卷叶螟危害的可行性,本研究于2021年6-10月开展了无人机对稻纵卷叶螟危害和田间水稻生育状况的同步观测试验,分析了15种植被指数与卷叶率(虫害指标)之间的相关关系;分别采用普通最小二乘法、多项式拟合、多元逐步回归法和偏最小二乘法建立了水稻分蘖期、拔节期和孕穗期的卷叶率反演模型;在此基础上,筛选出最优模型并分析卷叶率与水稻生理生态参数之间的关联。结果表明,(1)3个生育期的大部分植被指数与卷叶率存在极显著的相关性,每个生育期卷叶率与均一化植被指数(NDVI)的相关性都是最高的。(2)分蘖期的卷叶率反演模型效果最好,孕穗期的模型较好,拔节期的模型效果稍差。(3)在分蘖期,稻纵卷叶螟对水稻的危害反映在叶绿素浓度的下降和叶色的变化;在拔节期,虫害会引起水稻的补偿反应,导致叶绿素含量和叶面积增加;在孕穗期虫害对水稻生长的危害主要表现为叶绿素含量下降。本研究结果可为无人机遥感技术在区域范围内精确识别稻纵卷叶螟危害提供重要的参考。To explore the feasibility of using multispectral images taken by unmanned aerial vehicles(UAV)to monitor the damage of rice leaf folder(Cnaphalocrocis medinalis Güenée),a set of synchronous observation experiments on the damage of C.medinalis and rice growth status in paddy fields were conducted by using UAV,during the period from June to October in 2021.Correlations between 15 spectral vegetation indexes and leaf-roll rate(insect damage index)of C.medinalis were analyzed.Inversion models for the leaf-roll rates during the tillering stage,jointing stage and booting stage of rice were established by using the ordinary least square method,polynomial fitting,multiple stepwise linear regression method and partial least square method respectively.The optimal model was selected,and the relationships between leaf-roll rate and rice physiological and ecological parameters were analyzed.The results showed that,firstly,during the three growth stages,most of the vegetation indexes showed extremely significant correlations with the leaf-roll rate,and during each growth stage the correlation between leaf-roll rate and normalized difference vegetation index(NDVI)was the highest.Secondly,during the tillering stage,effect of the leaf-roll rate inversion model was the best,followed by models for the booting stage,but the model for the jointing stage was slightly worse.Thirdly,during the tillering stage,the damage of C.medinalis to rice reflected in decreasing of chlorophyll concentration and changing of leaf color.During the jointing stage,damage of C.medinalis might cause compensatory response in rice,resulting in increasing of chlorophyll content and leaf area.During the booting stage,the damage of C.medinalis to rice growth mainly presented as the decreasing of chlorophyll content.The research results can provide important reference for using UAV remote sensing technology to identify the damage of C.medinalis accurately in regional range.

关 键 词：稻纵卷叶螟 无人机遥感 植被指数 多光谱影像 反演模型 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]