Prediction of Moisture content in Maize leaves based on Hyperspectral Image

摘要

The moisture content of maize leaves under different moisture treatment is predicted by the hyperspectral imaging technique. The reflectance spectral of maize leaves is acquired in the spectral range of 900 to 1700nm by hyperspectral image system. Andthe moisture content of the same samples is acquired by traditional baking method. Then the partial least square regression (PLSR), MLR and BP model were used to analyze the correlation between hyperspectral data and moisture content. Good correlationswere found between moisture content of maize leaves and spectral information, the coefficient of determination is betterthan 0.85 with RMSEC and RMSEP less than 0.01. This work provides an effective method for prediction ofmoisture content of maize leaves by hyperspectral image technique.