基于无人机多光谱遥感的水稻株高估测方法  

Rice Plant Height Estimation Method Based on UAV Multispectral-remote Sensing

作　　者：刘建春 陈思 文波龙[2,4] 刘宏远[2,4] 李晓峰 LIU Jianchun;CHEN Si;WEN Bolong;LIU Hongyuan;LI Xiaofeng(College of Surveying and Prospecting Engineering,Jilin Architecture University,Changchun 130118,China;Northeast Institute of Geography and Agroecology,Chinese Academy of Sciences,Changchun 130102,China;Changchun Jingyuetan Remote Sensing Test Site,Chinese Academy of Sciences,Changchun 130102,China;Key Laboratory of Wetland Ecology and Environment,Northeast Institute of Geography and Agroecology,Chinese Academy of Sciences,Changchun 130102,China)

机构地区：[1]吉林建筑大学测绘与勘查工程学院,长春130118 [2]中国科学院东北地理与农业生态研究所,长春130102 [3]中国科学院长春净月潭遥感实验站,长春130102 [4]中国科学院东北地理与农业生态研究所湿地生态与环境重点实验室,长春130102

出　　处：《遥感信息》2023年第3期61-68,共8页Remote Sensing Information

基　　金：中国科学院战略先导专项课题(XDA28110502);吉林省教育厅科学技术研究项目(JJKH20220259KJ)。

摘　　要：针对数字表面模型计算水稻株高步骤繁琐、精度低的问题,提出了一种基于无人机多光谱植被指数的水稻株高估测模型。首先,在吉林省大安市水稻示范区采用无人机搭载一体式多光谱成像系统,获取地面采样距离为0.015 m的高清数字正射影像、数字表面模型和多光谱影像,并采集水稻实测株高;其次,基于多光谱影像数据提取反射率,计算了28种植被指数并根据他们与水稻株高的相关性进行优选,将优选后的植被指数与水稻株高利用不同的回归方法构建水稻株高估测模型并进行精度验证。结果表明:基于改良后冠层叶绿素含量指数MCCCI构建的幂函数水稻株高估测模型拟合效果最好,精度最高,R^(2)为0.876,RMSE为5.15 cm。所提出的无人机水稻株高估测方法为精准农业监测提供了新方向。Aiming at the problems of complicated steps and low accuracy of rice plant height estimation by digital surface model,a rice plant height calculation model based on UAV multispectral vegetation index is proposed.Firstly,in order to collect the ground truth rice plant height,a high-definition digital orthoimage,digital surface model,and multispectral image are first acquired in the rice demonstration area of Da’an city,Jilin province,using an unmanned aerial vehicle(UAV)outfitted with an integrated multispectral imaging system.The reflectance retrieved from the multispectral imaging data is used to produce the twenty-eight planting index,which is then optimized and selected according to their correlation with the relationship between the vegetation index and rice plant height.The height estimate model for rice plants is created and its correctness is checked using the different regression approaches.The analysis of the data reveals that the power function rice plant height estimate model,which is based on the modified canopy chlorophyll content index MCCCI,has the greatest fitting effect and the highest accuracy,with R^(2) of 0.876 and RMSE of 5.15 cm.Precision agricultural monitoring takes a new direction with the suggested UAV approach for estimating rice plant height.

关 键 词：植被指数 无人机 多光谱遥感 水稻株高 估测模型 

分 类 号：S252[农业科学—农业机械化工程]