论文部分内容阅读
采集农田、林地和盐碱地不同类型的土壤样本,采用偏最小二乘法结合OSC方法建立土壤有机质反演模型,运用交叉验证和外部验证相结合的评价方法进行比较分析。结果显示:采用平滑+MSC+OSC方法对光谱进行预处理,可以提高预测模型的精度。OSC因子个数和PLS主因子个数分别为6和4时,交叉验证决定系数R2为0.990 1,均方根误差为0.297 5,外部验证决定系数R2为0.926 1,均方根误差为0.283 6,模型达到最优。表明对光谱进行OSC预处理后建模是可行的,OSC降低与浓度阵无关的光谱信号,并且减少建立模型的主因子个数,进一步提高模型的精度和稳定性。
Soil samples collected from farmland, forestland and saline-alkali soil were collected. Partial least square method and OSC method were used to establish the soil organic matter retrieval model. The comparative analysis was made by cross-validation and external verification. The results show that using the smoothed + MSC + OSC method to preprocess the spectra can improve the accuracy of the prediction model. The number of OSC factors and the number of PLS main factors were 6 and 4 respectively, the cross validation coefficient R2 was 0.990 1, the root mean square error was 0.297 5, the external validation coefficient R2 was 0.926 1, and the root mean square error was 0.283 6 , The model to achieve the best. The results show that it is feasible to model the spectrum by OSC pretreatment. OSC can reduce the spectral signal which has nothing to do with the concentration array, and reduce the number of the main factors to establish the model to further improve the accuracy and stability of the model.