Fruit quality is important in different sections of fruit supply chain including harvest time, packaging, processing, grading, transportation, storage and shelf life. In the present study, the potential of Vis/SWNIR spectral region of 425–950 nm coupled with Chemometrics to predict anthocyanin, taste index (SSC/TA) and flesh firmness of fig fruit was investigated. Besides, the efficiency of ANN, KNN, and DA classifiers to classifying ripe, semi-ripe, and unripe figs were compared. A total of 167 fig fruits were used to calibrate models and validation. The regression ANN and classifiers (KNN, ANN, and DA) performances with common pretreatments (MA, SNV, MSC, etc.) and combinations of them were evaluated. Results showed that MA + 1st derivative + SG combined preprocessing, has highest mean value of RPD equal to 1.55 for flesh firmness prediction (RMSEP = 1.83, rp = 0.76) and model performance for anthocyanin and taste index predicting, wasn’t acceptable. ANN and DA provided mean classification accuracy of 89.86% and 89.52%, with MA+SNV and MA + MSC combined pretreatments, respectively. This study provides necessary information in the application of spectroscopy devices (425-950 nm) for the measurements of fig fruit quality attributes.
