基于无人机高光谱影像玉米叶绿素含量估算

Estimation of maize leaf chlorophyll contents based on UAV hyperspectral drone image

以无人机为平台搭载高光谱相机获得玉米农田高光谱影像,从中提取光谱特征参数,构建玉米叶片叶绿素含量估算模型,并制作玉米叶片叶绿素含量分布图。结果表明,以红边面积(SDr)、红边一阶微分最大值(Dr)、差值植被指数(DVI)为自变量构建的回归模型建模精度较高,以此反演玉米叶片SPAD值分布图并对填图结果进行精度检验,得出SPAD-Dr模型填图预测效果最佳(R2=0.89,RMSE=1.28,RE=2.31),可以作为玉米叶片叶绿素含量无人机高光谱影像遥感反演估算的基本模型。

UAV remote sensing system can quickly acquire high-resolution remote sensing images on farmland scale, which is significant for crop growth monitoring and agricultural production management. In this research, the hyperspectral images of maize field were acquired with a UHD185 camera mounted on a drone. The spectral parameters were extracted from the hyperspectral images to construct models for estimating chlorophyll content in maize leaves. Chlorophyll distribution maps of maize leaf were inversely estimated using these models. The results showed that the simple regression model separately built with red edge area ( SDr ), maximum first derivative values within red edge ( Dr ), or difference vegetation index ( DVI ) had higher modeling accuracy. These inversion models were used to make SPAD value distribution map of maize leaves, then, they were validated against observed results for the accuracy of the map, it was found that SPAD-Dr model was the best one in estimating the chlorophyll of maize leaves ( R 2 =0.89, RMSE =1.28, and RE =2.31). Therefore, this new method is feasible to be used for estimating the chlorophyll content of maize leaves.