Prediction of Moisture content in Maize leaves based on Near-infrared Spectroscopy

摘要

In this paper, we propose a novel measure method of the moisture content in maize leaves unde different moisture treatment by the near-infrared spectroscopy technique. The reflectance spectral of maizt leaves is acquired in the spectral range of 300 to 2500 nm by near-infrared imaging system. And the moistun content of the maize leaves is acquired by traditional baking method. Then the partial least square regressioi (PLSR) model was used to analyze the correlation between near-infrared data and moisture content measurei by tradition baking method. Good correlations were found between moisture content in maize leaves and near infrared spectral information. By our experimental results, the coefficient of determination is equal to 0.97 witl SEC of 0.004 and SEP of 0.0032. This work provides an effective method for prediction of moisture content o maize leaves by near-infrared spectral technique.