Rapid Nondestructive Postharvest Potato Freshness and Cultivar Discrimination Assessment

Background: Quality and safety of potato is both cultivar postharvest management dependent. The precise assessment freshness are complex tasks requiring time-consuming, expensive, destructive techniques. Method: Potatoes from three commercial cultivars were stored for 5 months at °C. Color chlorophyll fluorescence recorded, Red-Green-Blue (R-G-B), Red-Green-Near infrared (R-G-NIR) Red-Blue-Near (R-B-NIR) digital images, as well hyperspectral images acquired on the external periderm tuber in inner flesh part. Partial least square regression (PLSR) discriminant analysis, combined with feature selection techniques implemented, order to assess classify them into respective genotypes. Results: PLSR analysis visible/near (Vis/NIR) spectra reflectance most reliably predicted freshness, a cross-validated coefficient equal 0.981 0.947, determined by or internal measurements, respectively. Variance inflation factor, variable importance scores, genetic algorithms identified specific wavelength regions that mostly affected accuracy model terms strongest lowest collinearity root mean cross validation error. Conclusions: Vis/NIR data skin tubers may be used storage life, discrimination process.