|
|
|
| Estimation of chlorophyll content in rice canopy leaves based on main base analysis and dimensionality reduction method |
| YUAN Weinan1, XU Tongyu1,2*, CAO Yingli1,2, WANG Yang1, YU Fenghua1,2 |
(1. College of Information and Electrical Engineering, Shenyang Agricultural University, Shenyang 110161, China; 2. Liaoning Agricultural
Information Technology Center, Shenyang 110866, China) |
|
|
|
Abstract Due to high dimensional characteristics of unmanned aerial vehicle (UAV) hyperspectral remote sensing data, we proposed a dimension reduction method based on main base analysis. The 400-1 000 nm bands which are sensitive to chlorophyll, were selected to be subjected to Gram_Schmidt transform. After finding the projection space, the main base of concentrated band information was constructed, and the least square regression model was established to estimate the chlorophyll content in rice canopy leaves. The results showed that the modeling coefficient of determination (R2) was 0.689, with the root mean square error (RMSE) of 2.20, and the RMSE of validation model was 1.20. Compared with the prediction accuracy of the same model established by three vegetation indexes PRI, RD2 and MCARI, the modeling R2 was greatly improved, while the RMSE of verification model was greatly reduced. It is proved that the proposed method is effective, and it is significance for estimating chlorophyll content of plant leaves.
|
|
Published: 11 September 2018
|
|
|
基于主基底分析降维方法的水稻冠层叶片叶绿素含量估算
针对无人机高光谱遥感数据的高维特性，本文提出了一种基于主基底分析的降维方法。选取对叶绿素敏感的400～1 000 nm波段进行Gram_Schmidt 变换找到投影空间，构造集中波段信息的主基底，建立最小二乘回归模型来进行叶绿素含量估算。结果表明：基于主基底分析降维方法的建模决定系数（R2）为0.689，均方根误差（root mean square error, RMSE）为2.20，验证模型的RMSE为1.20；与3 种植被指数PRI、RD2 和MCARI降维后建立的相同模型预测精度相比，该方法的建模R2有了很大的提升，而验证模型的RMSE有所降低。研究结果验证了所提算法的有效性，对植物叶片的叶绿素含量估算具有重要意义。
关键词：
Gram_Schmidt变换,
波段降维,
主基底,
叶绿素,
无人机,
遥感
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
| |
Shared |
|
|
|
|
| |
Discussed |
|
|
|
|
