When the typical hyperspectral image classification method is used for vegetation classification and extraction, its extraction accuracy and precision classification degree are difficult to meet the practical application requirements. Hyperspectral data acquisition is mainly accomplished by two kinds of sensors: non-imaging spectrometer and imaging spectrometer. This paper combines SCVHR 1024i portable hyperspectral instrument and ubert UHD185 UAV imaging spectrometer to collect spectral reflectance curve of soybean, through data processing to achieve the purpose of analysis of its growth, and has a strong guiding significance for large-scale monitoring of soybean nutrition, pests, pesticide residues and so on.