Germination Prediction of Sugar Beet Seeds Based on HSI and SVM-RBF

Beet is an important sugar crop in China. The selection of beet seeds is a key link in the process of agricultural breeding. Hyperspectral technology has the advantages of fast, real-time, accurate and lossless acquisition of seed morphological characteristics, internal structural characteristics, chemical composition and other characteristic information, and has a good application prospect in seed quality testing, classification and identification. In this study, the near infrared hyperspectral image acquisition system was used to obtain the hyperspectral images of 3072 samples. The average spectrum of seed area was extracted as its characteristic spectrum. Ten characteristic wavelengths of characteristic spectrum were selected by continuous projection algorithm, and then the model was established by SVM-RBF algorithm. The model accuracy of this test device is 87.3%. The results show that high spectral imaging can predict the germination of beet seeds accurately, which provides a new idea for online nondestructive testing of beet seeds.