Study on Chlorophyll Content in Visible-Near Infrared Spectroscopy

Chlorophyll, as an important factor for the normal growth and development of plants, is of great significance for the management of agricultural water and fertilizer. In this study, the chlorophyll content of maize leaves was taken as the research object, and the chlorophyll content prediction model was quantitatively studied. ASD FieldSpec Pro spectrometer was used to measure the spectral reflectance of the leaf samples, and the spectral curve characteristics of different contents were analyzed. The results show that: (1) The reflection spectrum undergoes the nine-point smoothing of Savitzky-Golay and combines with the MSC, NOR, and SNV transforms to significantly increase the signal-to-noise ratio of the reflection spectrum. The combination band with higher correlation can significantly improve the stability and prediction of the model ability. (2) In the PLSR model, MSC processing is performed on the smoothed spectrum, and the model established after the second-order differential transformation has the best effect, Rc(2) = 0.95, RMSEC = 2.32, SEC = 1.35.