Method for predicting content of flavonoid substances in wetted pine needle leaves

Abstract

The invention discloses a method for predicting the content of flavonoid substances in wetted pine needle leaves. The method comprises the following steps: collecting near infrared spectrum data of a wet and pine needle leaf calibration set sample, and carrying out spectrum pretreatment; measuring the flavonoid content of the needle leaves in the calibration set sample by using a conventional method; correlating the preprocessed spectral data with the measured flavonoid substance content, performing regression analysis by using a support vector SVR, establishing a correction model, and performing optimization; and during prediction, scanning a sample to be detected by using a near infrared spectrum, and substituting spectral characteristics into the model to obtain a predicted value of the flavonoid content of the wet pine needle leaves. The method provided by the invention can realize rapid, accurate and lossless prediction of the content of the flavonoids in the pine needle leaves in the breeding process of the raw material pines of the pines, and provides a solid scientific basis for screening high-yield and high-quality germplasm resources of the flavonoids of the pines of the pines; and important reference value is provided for building a raw material forest for high-yield and high-quality flavonoid substances.

Images