多植被指数组合的冬小麦遥感估产方法研究

发布时间：2018-03-09 23:34

本文选题：遥感估产　切入点：植被指数组合　出处：《干旱区资源与环境》2017年07期 　论文类型：期刊论文

【摘要】：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。
[Abstract]:In order to improve the accuracy of fast estimation of winter wheat yield in a large area of winter wheat, Landsat 8 OLI satellite remote sensing data were selected. The normalized vegetation index (NDVI), the ratio vegetation index (RVI), the green vegetation index (GVI) and the enhanced vegetation index (EVI) were calculated. The regression equations or neural networks and SVM models were established between the four cropping index combinations and the measured yield of winter wheat on the ground, respectively. The results show that the yield estimation accuracy of nonlinear regression equation of single vegetation index is higher than that of linear regression equation. The linear regression equation showed that the combination accuracy of multi-vegetation index was higher than that of single vegetation index, and the combination of multi-vegetation index could achieve complementary information and improve the precision of remote sensing yield estimation. The root mean square error (RMSE) of the model is 339.6 kg 路hm ~ (-1) ~ (-2), and the coefficient of determination is 0.7852, which is higher than that of BP neural network model. It can be applied to the rapid and accurate estimation of winter wheat yield by remote sensing.
【作者单位】：辽宁师范大学城市与环境学院;广西师范大学计算机科学与信息工程学院;北京师范大学地理学与遥感科学学院遥感科学国家重点实验室;
【基金】：国家自然科学基金项目(41561008)资助
【分类号】：S127;S512.11

【相似文献】