基于高光谱数据的晚籼稻品种剑叶SPAD值估测

Estimation of the Flag Leaf SPAD Value in Different Late Indica Rice Varieties Based on Hyperspectral Data

本研究测定了长江中下游不同晚籼稻品种的剑叶SPAD值及反射光谱数据,分析了原始光谱及其变换数据与SPAD值的相关关系,建立了不同晚籼稻品种剑叶SPAD值估测模型,并采用平均偏差率对模型精度进行品种间验证。结果表明,水稻剑叶SPAD值与原始光谱的敏感波段位于710~720 nm之间,与一阶微分光谱的敏感波段为690~700 nm,各光谱参数中以红边幅值(Dr)的相关性最好,相关系数达0.6~0.8;在回归模型类型的选取上,指数模型和二次曲线模型的估测效果较优;715 nm处的原始光谱反射率建立的指数模型y=85.512e~(-2.392x)的估测效果最好,进行品种间验证的平均偏差率最小,仅为5.01%。说明以单一晚籼稻品种叶片光谱参数建立的模型在品种间具有一定的适用性。

In this study,the SPAD values and reflectance spectra of flag leaves of different late indica rice varieties in the middle and lower reaches of the Yangtze river were measured.The estimation models of SPAD values of flag leaves in different late indica rice varieties was established by analyzing the correlation between the original spectral reflectance and its transformed data and SPAD values.And the average deviation rate was used to verify the model accuracy.The results showed that the sensitive band of SPAD value and the original spectral reflectance of rice leaves was between 710 nm and 720 nm,and the sensitive band of the first derivative spectrum was approximately 690 nm to 700 nm.Among all spectral parameters,the red edge value(Dr)had the best correlation,with the correlation coefficient of 0.6~0.8;the exponential model and quadratic curve model have better estimation effect in the selection of regression model;the index model y=85.512e-2.392x based on the original spectral reflectance at 715 nm had the best estimation effect,and the average deviation rate for inter-species verification was the lowest(5.01%).The results showed that the model established with the spectral parameters of the leaves of single late indica rice variety had certain applicability among different varieties.