Application of NIR Spectroscopy in Classification of Plant Species

Near-infrared spectroscopy (NIRS) analysis is a rapidly-developing non-destructive technology with increasingly wide range of applications, but application of NIRS technology in the classification of plants is still rare. Based on experimental study, this paper uses the approach of principal component analysis (PCA) combined with the establishment of a Mahalanobis distance of plants’ leaves to establish the discrimination model of leaves, and adopts multiplicative scatter correction (MSC) and a first-order differential to analyze the model. The analysis results are quite consistent with experimental data, and obtained a satisfactory classification rate of 100%, which shows that this approach can be used to classify plant species, and it can become a very useful tool in many relevant fields.