Non-destructive quality determination of frozen food using NIR spectroscopy-based machine learning and predictive modelling

Non-destructive testing for the quality of frozen food is of great interest. A model food product was developed as the test material for this study. Different modeling methods were applied to establish the relationship between the near-infrared (NIR) spectra of the frozen samples and quality indicators of drip loss, texture parameters including hardness, chewiness, gumminess and gel strength, respectively. Principal component analysis (PCA) and hierarchical clustering analysis (HCA) analysis results show that the collected NIR spectra of the model food prepared based on different moisture content were well distinguished. The modeling results show that principal component regression (PCR), support vector machine regression (SVR), partial least squares regression (PLSR) and back-propagation artificial neural network (BP-ANN) algorithms could be used to predict the quality indicators of frozen samples. By comparison, the BP-ANN modeling approach performed better with higher R2 and lower root mean squared errors (RMSE).