Modelling postharvest quality of blueberry affected by biological variability using image and spectral data

BACKGROUND Hyperspectral reflectance and transmittance sensing as well as near‐infrared ( NIR ) spectroscopy were investigated as non‐destructive tools for estimating blueberry firmness, elastic modulus and soluble solid content ( SSC ). Least squares–support vector machine models were established from these three spectra based on samples from three cultivars viz. Bluecrop , Duke and M2 and two harvest years viz. 2014 and 2015 for predicting blueberry postharvest quality. RESULTS One‐cultivar reflectance models (establishing model using one cultivar) derived better results than the corresponding transmittance and NIR models for predicting blueberry firmness with few cultivar effects. Two‐cultivar NIR models (establishing model using two cultivars) proved to be suitable for estimating blueberry SSC with correlations over 0.83. R p ( RMSE p ) values of the three‐cultivar reflectance models (establishing model using 75% of three cultivars) were 0.73 (0.094) and 0.73 (0.186), respectively , for predicting blueberry firmness and elastic modulus. For SSC prediction, the three‐cultivar NIR model was found to achieve an R p ( RMSE p ) value of 0.85 (0.090). Adding Bluecrop samples harvested in 2014 could enhance the three‐cultivar model robustness for firmness and elastic modulus. CONCLUSION The above results indicated the potential for using spatial and spectral techniques to develop robust models for predicting blueberry postharvest quality containing biological variability. © 2015 Society of Chemical Industry