Nondestructive Testing Model of Mango Dry Matter Based on Fluorescence Hyperspectral Imaging Technology

The dry matter test of mango has important practical significance for the quality classification mango. Most common fruit and vegetable nondestructive testing methods based on fluorescence hyperspectral imaging technology use a single algorithm in algorithms such as Uninformative Variable Elimination (UVE), Random Frog (RF), Competitive Adaptive Reweighted Sampling (CARS) Continuous Projection Algorithm (SPA) to extract feature spectral variables, these alone can easily lead insufficient stability prediction results. In this regard, detection method was carried out. Taking ‘Keitt’ research object, samples were numbered sequence, their images wavelength range 350–1100 nm collected, average spectrum region interest used effective information sample. Select SPXY divide into calibration set set, select Orthogonal Signal Correction (OSC) preprocessing method. For preprocessed spectra, primary dimensionality reduction (UVE, SPA, RF, CARS), combined (UVE + CARS SPA), secondary ((CARS SPA)-SPA, RF)-SPA) other 12 variables. Separately constructed predictive models predicting mangoes, namely, Support Vector Regression (SVR), Extreme Learning Machine (ELM), Back Propagation Neural Network (BPNN) model, used; results show that (CARS RF)-SPA-BPNN best performance matter, its correlation coefficients RC2 = 0.9710, RP2 0.9658, RMSEC 0.1418, RMSEP 0.1526, provides reliable theoretical basis technical support non-destructive detection, precise intelligent development detection.