Detection of Moisture Content of Pinus massoniana Lamb. Seedling Leaf Based on NIR Spectroscopy with a Multi-Learner Model

The growth quality of Pinus massoniana (Lamb.) seedlings is closely related to the survival rate afforestation. Moisture content detection an important indicator in cultivation forest because it can directly reflect adaptability and potential soil environment. To improve accuracy quantitative analysis moisture P. using near-infrared spectroscopy, a total 100 were collected, their diffuse reflectance spectra measured range 2500 800 nm (12,000 4000 cm−1). An integrated learning framework was introduced, model for established by combining preprocessing feature wavelength selection methods chemometrics. Our results showed that information carried after multiple scattering correction (MSC) had good response target attribute. stacking based on full-band spectrum prediction coefficient determination R2 0.8819, could be significantly improved compared single model. After variable selection, processed MSC with uninformative elimination (UVE) effects all models. Additionally, support vector regression (SVR)—adaptive boosting (AdaBoost)—partial least squares (PLSR) + AdaBoost reached 0.9430. This indicates through preprocessing, models achieve high predicting seedlings. provide feasible technical reference precision