基于高光谱数据的苔草营养成分反演方法研究

Inversion Method of Carex Nutritional Ingredient Based on Hyperspectral Remote Sensing Data

研究基于高光谱数据的苔草营养成分(侧重粗蛋白质、总氮、总磷)反演方法。结果显示,粗蛋白质的最佳反演模型是通过原始光谱反射率(偏最小二乘回归的方法)获得,R2=0.814、RMSE=0.450;总氮的最佳反演模型是通过一阶光谱反射率(偏最小二乘回归的方法)获得,R^2=0.850、RMSE=0.175;总磷的最佳反演模型是通过原始光谱反射率(偏最小二乘回归)获得,R^2=0.882、RMSE=0.025。最佳模型检验结果显示估算值和实测值之间的强相关性:粗蛋白质R2=0.801、RMSE=1.029,总氮R2=0.777、RMSE=0.234,总磷R2=0.756、RMSE=0.043。

This paper researched the inversion method of carex nutritional ingredient based on hyperspectral remote sensing data. The results show that the optimal inversion model for crude protein is obtained using partial least squares regression method(PLS) and the original spectra reflectance(R^2=0.814, RMSE=0.450), the optimal one for total nitrogen(TN) is gotten using PLS and the first derivative spectra reflectance(R^2=0.850, RMSE=0.175), and the optimal one for total phosphorus(TP) is derived using PLS and the original spectra reflectance(R^2=0.882, RMSE=0.025). The validation results of these optimal models show strong correlations between estimated and measured values, such as PT with R^2=0.801 and RMSE=1.029, TN with R^2=0.777 and RMSE=0.234, and TP with R^2=0.756 and RMSE=0.043.