基于叶片反射光谱估测水稻氮营养指数
|
徐浩聪,姚波,王权,陈婷婷,朱铁忠,何海兵,柯健,尤翠翠,吴小文,郭爽爽,武立权
|
Determination of Suitable Band Width for Estimating Rice Nitrogen Nutrition Index Based on Leaf Reflectance Spectra
|
XU HaoCong,YAO Bo,WANG Quan,CHEN TingTing,ZHU TieZhong,HE HaiBing,KE Jian,YOU CuiCui,WU XiaoWen,GUO ShuangShuang,WU LiQuan
|
|
表6 水稻NNI(y)与不同光谱指数(x)的定量关系(n=160)及模型检验效果(n=249)
|
Table 6 Quantitative relationships between NNI (y) and different spectral index (x) in rice (n=249)
|
|
叶位 Leaf position | 光谱指数 Spectral index | 回归方程 Regression equation | 模型精度 Model precision (R2) | 预测精度 Prediction precision (R2) | 均方根误 RMSE | 相对误差 RE (%) | 顶一叶(L1) | SR(R900,R540) | y=0.2488x+0.1267 | 0.390 | 0.443 | 0.187 | 14.7 | SR[AR(900±50),AR(540±10)] | y=0.2454x+0.1307 | 0.388 | 0.441 | 0.187 | 14.7 | 顶二叶(L2) | SR(R900,R540) | y=0.3943x-0.3199 | 0.657 | 0.707 | 0.136 | 12.2 | SR[AR(900±50),AR(540±10)] | y=0.3896x-0.3154 | 0.654 | 0.706 | 0.136 | 12.2 | 顶三叶(L2) | SR(R900,R540) | y=0.3019x-0.0364 | 0.557 | 0.731 | 0.130 | 11.6 | SR[AR(900±50),AR(540±10)] | y=0.2988x-0.0344 | 0.556 | 0.729 | 0.130 | 11.6 | 顶二、顶三光谱平均(L23) | SR(R900,R540) | y=0.3556x-0.1997 | 0.625 | 0.740 | 0.128 | 11.5 | SR[AR(900±50),AR(540±10)] | y=0.3517x-0.1967 | 0.624 | 0.740 | 0.128 | 11.5 | 顶一、顶二、顶三光谱平均(L123) | SR(R900,R540) | y=0.3645x-0.2208 | 0.618 | 0.720 | 0.133 | 11.2 | SR[AR(900±50),AR(540±10)] | y=0.3602x-0.217 | 0.616 | 0.718 | 0.133 | 11.3 |
|
|
|