为推进高光谱遥感技术在烟草病害监测和病害预测上的应用,利用野外光谱仪在山东省费县烤烟种植区研究了发病烤烟叶片与健康烤烟叶片的高光谱反射率特征。研究发现,烤烟病叶与健康叶片的光谱反射率有明显差异,病叶的高光谱反射率明显高于健康叶片;在550 nm处,随着叶片发病程度的增加,绿峰有向红光方向移动的趋势。随着病情程度的增加,烟草单叶光谱的红边位置向短波方向移动,即红边斜率减小,红边发生"蓝移";病害越重,蓝移越明显。利用高光谱遥感技术,可以预测和监控烤烟病害的发生及蔓延。
To promote the application of hyperspectral remote sensing in monitoring and predicting the diseases of flue-cured tobacco, hyperspectral reflectance of infected and healthy tobacco leaves were studied by using ASD hand-held spectroradiometers in Fei County of Shandong province. Results suggested that there was remarkable difference in hyperspectral reflectance between infected tobacco leaves and healthy ones, and reflectivity of infected tobacco leaves was higher than that of healthy ones; as tobacco disease grew worse, green peak(550nm) of reflection spectrum moved to long wave. First derivative of reflectance curve of single tobacco leaf told us that "red edge" of reflection spectrum of infected leaves moved to short wave, which became more obvious when tobacco disease was getting worse. So it was proved that tobacco diseases could be monitored and predicted by hyperspectral remote sensing.