剔除土壤背景对反演玉米根域土壤含水率的影响研究

Effect of Removing Soil Background on Inversion of Soil Moisture Content in Maize Root Zone

为研究剔除土壤背景对无人机多光谱监测土壤含水率的影响,通过无人机飞行拍摄得到多光谱影像数据,运用监督分类剔除土壤背景,并分别提取剔除背景前后各6个波段的光谱反射率。将反射率与深度10 cm、20 cm和30 cm的实测含水率分别构建一元线性回归模型、逐步回归模型、偏最小二乘回归模型和岭回归模型,以R^(2)、RMSE、RE为指标进行模型精度评价。结果表明:剔除灌浆期玉米的土壤背景后,数据的相关性和回归模型的精度始终比未剔除土壤背景数据的相关性和模型精度差。剔除土壤背景的反射率与各个深度含水率的相关系数绝对值在0.01~0.33之间,未剔除土壤背景的反射率与各个深度含水率的相关系数绝对值在0.08~0.54之间。未剔除土壤背景模型的效果在任何深度均高于剔除土壤背景的模型,其中在各深度处偏最小二乘回归模型的精度均为最优。

In order to study the influence of removing the soil background on the unmanned aerial vehicles(UAV)multi-spectral monitoring of soil moisture content,the multi-spectral image data was obtained through UAV flight shooting,the soil background was removed by supervised classification,and the spectral reflectance of 6 bands before and after the background removal were extracted.The reflectance and the measured water content at depths of 10 cm,20 cm and 30 cm were used to construct univariate linear regression model,stepwise regression model,apartial least-squares regression model,and ridge regression model,respectively.R^(2),RMSE,and RE were used as indicators to evaluate the accuracy of the model.The results showed that the correlation of the data and the accuracy of the regression model were always worse than that after excluding the soil background when the maize at the filling stage.The absolute value of the correlation coefficient between the reflectance of the soil background and the moisture content of each depth ranges from 0.01 to 0.33,and the absolute value of the correlation coefficient of the reflectance of the soil background and the moisture content of each depth ranges from 0.08 to 0.54.The effect of the model with soil background is higher at any depth than that of the model without soil background,and the accuracy of the partial least-squares regression model is the best at each depth.