论文部分内容阅读
为提高高光谱植被指数对棉花叶绿素含量的估算精度,以陕西省关中地区棉花花铃期叶片高光谱反射率为数据源,分析了13种植被指数与棉花叶片叶绿素相对含量(SPAD)的相关关系;同时采用降精细采样法,详细分析400~2 000nm波段范围内原始光谱反射率的任意两两波段组合而成的优化光谱指数RSI与SPAD值的定量关系,构建线性及非线性回归监测模型,并对模型进行验证。结果表明:1)所提取的13种植被指数中NIR/NIR与SPAD值的相关系数最大(r=0.914),并且基于NIR/NIR(R780/R740)构建的回归方程模型优于其他植被指数,其构建的二次曲线方程回归模型建模与验模R2分别为0.900和0.785,RMSE为4.762,RE为7.86%,为基于提取的12种植被指数构建SPAD值估算模型中最佳模型;2)优化后的比值光谱指数RSI(Ration spectral index)的敏感波段为500和563nm,RSI(500,563)与SPAD值的相关系数r=0.999,与棉花叶片SPAD含量在0.01水平下呈显著相关,其构建的二次曲线方程模型效果最优,建模和验模R2分别为0.912和1.000,RMSE为2.848,RE为4.38%。与提取的13种植被指数相比,基于RSI指数二次曲线回归模型为估算叶绿素含量的最佳模型,并且模型预测值和实测值之间的符合度较高R2=0.843,表明基于波段优化算法的优化光谱指数RSI能更好的预测棉花叶片叶绿素含量。
In order to improve the accuracy of estimating the chlorophyll content of cotton by hyperspectral vegetation index (NDVI), the correlation between the 13 vegetation indices and the chlorophyll content (SPAD) of cotton leaves was analyzed based on the hyperspectral reflectance at the boll stage of cotton in Guanzhong, Shaanxi Province. At the same time, by using the down-fine sampling method, the quantitative relationship between the optimal spectral index (RSI) and the SPAD value, which is a combination of any two bands of the original spectral reflectance in the 400-2000 nm band, was analyzed in detail to establish a linear and nonlinear regression monitoring model, And verify the model. The results showed that: 1) the correlation coefficient between NIR / NIR and SPAD was the highest (r = 0.914), and the regression equation model based on NIR / NIR (R780 / R740) was superior to other vegetation indices, The regression model modeling and model R2 of the constructed quadratic equation were 0.900 and 0.785 respectively, RMSE was 4.762 and RE was 7.86%, which was the best model for estimating the SPAD value based on the extracted 12 vegetation indices. 2) The sensitivity bands of the optimized Ration Spectral Index (RSI) were 500 and 563 nm, and the correlation coefficient of RSI (500, 563) and SPAD was 0.999, which was significantly correlated with SPAD content of cotton leaves at 0.01 level. The quadratic curve model works best with R2 of 0.912 and 1.000 respectively for modeling and validation, RMSE of 2.848, and RE of 4.38%. Compared with the 13 vegetation indices extracted, RSI index conic regression model was the best model for estimating chlorophyll content, and the coincidence of model predictive value and measured value was higher R2 = 0.843, indicating that based on band optimization The optimized spectral index RSI of the algorithm can better predict the chlorophyll content of cotton leaves.