Detection of the Contents of the Ingredient of Food by Using the NIR Spectroscopy and the Backward Interval Partial Least-Squares

Abstract

Near-infrared spectroscopy (NIRS), with the characteristics of high speed, non-destructiveness, high precision and reliable detection data etc., is a pollution-free, rapid, quantitative and qualitative analysis method. A new approach for the discrimination of the ingredients of corn (moisture, oil, protein, starch) by means of NIR spectroscopy (1100-2498 nm) was developed in this work. The relationship between the reflectance spectra and the ingredients of corn was established. The data were spilt into training and testing subsets by sample set partitioning based on join x-y distance (SPXY),the spectral data was compressed by orthogonal signal correction (OSC), wavelength was selected by backward interval partial least-squares (biPLS),the 60 samples to build PLS mode, the model was used to predict the varieties of 20 unknown samples. The standard error of prediction (SEP) was 0.173; the relative error of prediction (PRE) was 0.55%; the correlation coefficient (R) was 0.98. The way to detect the ingredient of food is simply, reliable.