*2.6. Establishing Chlorophyll Content Detection Model Based on PLSR*

PLSR gradually became a widely used modeling method in spectral analysis after it was proposed by Geladi in 1986 [48–50]. PLSR can solve the problems of autocorrelation and multicollinearity between variables on the basis of the method of principal component extraction. PLSR was used to perform principal component decomposition simultaneously on the spectral reflectance matrix and LCC matrix. During decomposition, PLSR correlated the spectral and chlorophyll content matrixes and established a linear regression model between the two to detect the chlorophyll content of corn leaves. The leave-one-out cross-validation (LOOCV) method was used for internal interactive verification, and the optimal number of characteristic variables was determined by root-mean-square error of cross-validation (RMSECV). The model evaluation indicators were the validation coefficient of validation set model (*Rv* 2) and the root-mean-square error of validation set (*RMSEV*).
