论文部分内容阅读
为了提高大面积冬小麦农田产量快速估算的准确率,选取Landsat 8 OLI卫星遥感数据,计算归一化植被指数NDVI、比值植被指数RVI、绿度植被指数GVI、增强植被指数EVI,分别建立4种植被指数组合与地面冬小麦实测产量的回归方程或神经网络和SVM模型。结果表明:单植被指数的非线性回归方程估产精度高于线性回归方程,冬小麦实测产量与遥感植被指数表现为非线性关系;线性回归方程估产时多植被指数组合精度高于单植被指数,多植被指数组合可实现信息互补,提高遥感估产精度;建立多植被指数组合与实测产量的非线性遥感估产模型时,SVM模型的均方根误差RMSE为339.6kg·hm~(-2),决定系数R~2为0.7852,估产精度高于BP神经网络模型、RBF神经网络模型,可应用于冬小麦遥感估产的快速、准确实现。
Landsat 8 OLI satellite remote sensing data was used to calculate the NDVI, RVI, GVI and EVI to improve the accuracy of large-scale winter wheat crop yield estimation. Four vegetation types Regression equation or neural network and SVM model of exponential combination with measured winter wheat yield. The results showed that the estimation accuracy of nonlinear regression equation of single-vegetation index was higher than that of linear regression equation, the measured yield of winter wheat showed a nonlinear relationship with remote sensing vegetation index, and the linear regression equation was more accurate than single-vegetation index Index combination can realize information complementarity and improve the accuracy of remote sensing estimation. When establishing a nonlinear remote sensing yield model with multi-vegetation index combination and measured yield, the root mean square error (RMSE) of SVM is 339.6 kg · hm -2 and the coefficient of determination R ~ 2 is 0.7852, and the estimation accuracy is higher than BP neural network model and RBF neural network model, which can be applied to the fast and accurate estimation of winter wheat yield.