论文部分内容阅读
人工田间会诱发不同等级的小麦条锈病,在不同生育期需测定染病冬小麦冠层光谱以及相应小麦的病情指数。把冠层光谱一阶微分数据与相应的小麦病情指数进行相关分析,采用单变量线性和非线性回归技术,选取部分样本建立小麦的病情指数估测模型,并利用其余的样本对模型进行检验。结果表明,病情指数与一阶微分在432~582nm,637~701nm和715~765nm波长区域内具有极显著的相关性。以蓝边内一阶微分总和(SDb)与红边内一阶微分总和(SDr)的归一化值作为变量的模型是估测病情指数的最佳模型,其RMSE为5.73%。研究表明,可用高光谱信息监测作物的病害情况,且精度较高。利用高光谱遥感监测病害程度及其影响具有实际的应用价值。
Artificial field will induce different levels of wheat stripe rust, at different growth stages to determine the winter wheat canopy disease spectrum and the corresponding wheat disease index. Correlation analysis between the first derivative data of canopy spectrum and the corresponding wheat disease index was carried out. Using univariate linear and nonlinear regression techniques, some samples were selected to establish the disease index estimation model of wheat and the remaining samples were used to test the model. The results showed that the disease index and first-order differential in the 432 ~ 582nm, 637 ~ 701nm and 715 ~ 765nm wavelength region has a very significant correlation. The model with the normalized values of the first derivative sum (SDb) in the blue margin and the first derivative sum (SDr) in the red edge as the variables is the best model to estimate the disease index with an RMSE of 5.73%. Studies have shown that hyperspectral information can be used to monitor crop disease and the accuracy is high. The use of hyperspectral remote sensing to monitor disease severity and its impact has practical value.