基于不同玉米品种叶片高光谱的SPAD值估测模型研究*

通过研究不同玉米品种叶片SPAD值与高光谱参数的关系,建立玉米叶片SPAD值估测模型,并对模型进行品种间精度检验,以期得到普遍适用且精度更高的模型。本研究通过两年试验,测定不同玉米品种的叶片SPAD值及其高光谱数据,综合分析叶片SPAD值与高光谱反射率、反射率一阶导数及其光谱参数的相关关系,对玉米叶片SPAD值估测模型进行构建。玉米叶片SPAD值与高光谱反射率最敏感波段在550nm和710nm附近,与反射率一阶导数最敏感波段出现在500nm~750nm范围内。叶片SPAD值与单波段反射率的相关性要高于其一阶导数,以550nm附近光谱反射率构建的模型对大多数品种的叶片SPAD预测值平均误差最小。 英文摘要: By studying the relationship between the leaf SPAD values of different kinds of maize and high spectral parameters, estimate models of the leaf SPAD values in maize were established and accuracy tests among varieties were done to get generally applicable and higher precision models. Through 2 years’ experiments, the leaf’s SPAD value and the hyperspectral data of different kinds of maize were measured, the experiment also comprehensively analyzed the leaf’s SPAD value, high spectral reflectivity, reflectivity first derivative and its correlation relationship of spectrum parameters, and the estimating model of the leaf’s SPAD value in maize was built. The result showed that the most sensitive waveband of the leaf’s SPAD value in maize and hyperspectral reflectivity existed in the near 550 nm and 710 nm, and between the leaf’s SPAD value in maize and the first derivative of reflectivity, the most sensitive waveband were in the range of 500 nm to 750 nm. The correlation of leaf’s SPAD value and the single waveband reflectivity was higher than its first derivative, and the precision of the model which was based on the spectral parameters 550nm relative error on average was least.
;