Non-destructive determination of internal defects in chestnut (Castanea mollissima) during postharvest storage using X-ray computed tomography

After harvest at horticultural maturity, X-ray computed tomography was used for the non-destructive determi-nation of internal defects occurring in chestnuts during storage. An image processing pipeline was established to accurately identify various types of defective chestnuts. Severely defective chestnuts, with a complete loss of eating quality, were accurately identified by their gray scale value from volume and porosity analyses. Slightly defective calcified chestnuts were classified by local voxel gray scale value analysis. Minor isolated defects in chestnuts (defects only in the kernel) were distinguished by morphological analysis. An internal pore network model was also used to effectively identify connected defective chestnuts (defects connected to the inner skin), with the type of defect being distinguished by spatial distribution morphology. Overall, X-ray computed to-mography provided a new perspective for the non-destructive characterization of the three-dimensional image structure of chestnuts and provided an important concept for screening and classifying their storage lesions.