基于光谱反射特征的柑橘叶片含水率模型

采集了46个柑橘叶片信息，其中一半用于模型构建，一半用于模型验证，本文采用光谱逐步回归分析和构造光谱指数两种方法分析了叶片光谱（380-2500nm）反射率与含水量之间的关系。构建了柑橘叶片含水量光谱反射率模型，发现光谱逐步回归分析所得模型精度高于后者，但从物理含义和逻辑性方面考虑，推荐光谱指数法构建的模型。

Model establishment for citrus leaves dry-basis moisture content based on spectral signature

In the experiment, data of 46 citrus leaves was collected, half of them were used to establish model, and other half of them were used to verify model. The relationship between leaf water content and leaf reflectance was studied, models of leaves dry-basis moisture content based on spectral were established using stepwise regression method and constructing spectral indexes method respectively. As the result, model established using stepwise regression method has higher precision than model using another method. But we recommend model established using constructing spectral indexes method considering the physical meaning.