Non-destructive porosity mapping of fruit and vegetables using X-ray CT

•3-D porosity maps of horticultural products were created from X-ray CT.•A single correlation model was applicable to different products.•The method was successfully validated for products with different porosities.•A large heterogeneity of porosity was found in fruits and vegetables. Fruit and vege...

Full description

Saved in:
Bibliographic Details
Published in:Postharvest biology and technology 2019-04, Vol.150, p.80-88
Main Authors: Nugraha, Bayu, Verboven, Pieter, Janssen, Siem, Wang, Zi, Nicolaï, Bart M.
Format: Article
Language:English
Subjects:
3-D imaging
Apple
Eggplant
Pear
Turnip
X-ray computed tomography
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•3-D porosity maps of horticultural products were created from X-ray CT.•A single correlation model was applicable to different products.•The method was successfully validated for products with different porosities.•A large heterogeneity of porosity was found in fruits and vegetables. Fruit and vegetables have a considerable variability in porosity affecting transfer phenomena caused by respiration during postharvest storage. In this study, a validated and reproducible method to calculate and map the porosity distribution in fruit and vegetable organs based on a grayscale-porosity correlation model is introduced. The method requires a water phantom for proper evaluation of X-ray computed tomography (CT) images and juices of scanned plant products for 0% porosity references. A strong correlation was found between grayscale values and porosity (R2 = 0.99) enabling to estimate the complete 3-D porosity distribution of eggplant, turnip, apple and pear using a single correlation function. The porosity maps well depicted the heterogeneity of tissue microstructure either in the individually intact product or among the samples. Porosity differences in structural details such as different tissues, seeds and vascular bundles could be clearly distinguished. In average, eggplant exhibited the most porous structure (41.8 ± 1.0%) followed by turnip (23.3 ± 3.4%), apple (19.7 ± 1.1%) and pear (4.0 ± 1.6%). The developed method allows to predict porosity distribution of other plant products non-destructively by only including juice scans. A normalization procedure make the method effortlessly adaptable to other CT systems and products.
ISSN:0925-5214
1873-2356
DOI:10.1016/j.postharvbio.2018.12.016