Apple Multi-Index Fusion Classification Based on Near Infrared Spectroscopy

Apples' soluble solids content (SSC) and moisture content (MC) are the key factors for affecting the taste of apple. Non-destructive testing (NDT) and post-harvest grading of apples are of great significance to increase the economic value and enhance the market competitiveness of apples. At present, the research on apples' NDT is mainly aimed at single index (such as SSC, MC, etc.), which can not fully reflect the internal quality. Therefore, based on near infrared spectroscopy (NIR), this paper uses long and short-term memory neural network (LSTM) method to establish a prediction model to achieve the classification forecast apples' multi-index, which can reflect the comprehensive quality of apples more reasonably.