Classification of jaboticaba fruits at three maturity stages using NIRS and LDA

摘要: This study proposes a rapid and non-destructive method of jaboticaba [Myrciaria cauliflora (Mart.) O. Berg] fruit classification at three maturity stages based on the skin colour (immature – completely green, physiologically mature turning from green to purple, ripe purple) using Near-Infrared Reflectance Spectroscopy (NIRS) combined with principal component analysis-linear discriminant analysis (PCA-LDA), variable selection techniques employing successive projection algorithm (SPA-LDA) or genetic (GA-LDA). One hundred eighty samples in were used multivariate accuracy results tested sensitivity, specificity, positive (or precision) negative predictive values, Youden index, likelihood ratios. In immature stage models PCA-LDA, GA-LDA SPA-LDA achieved sensitivity 100% validation set. The obtained this suggest that proposed is promising alternative for assessing maturity, opening possibility automation packing houses.