On the use of a non-destructive monitoring technique to evaluate the quality loss of strawberries during microwave drying processing

GAMBOA SANTOS J.; VASCO, M. F.; CAMPAÑONE, L.A.

Amsterdan

Conferencia; 15th International Conference on Heat Transfer, Fluid Mechanics and Thermodynamics (ATE-HEFAT); 2021

Elsevier

This work aims to analyse the potential of digital image analysis (DIA) to extract valuable information about optical and morphological properties of fresh (FR) and pre-treated (osmo-dehydrated, OD; sucrose, 60 ºB, 40 ºC, 4h) strawberries during a microwave (MW) assisted drying (1.2W/g, 100 min) process. 1150 images of individual strawberry samples, 490 for FR and 660 for OD samples, were analysed. Differences among sample type (FR and OD) and time categories (short, 70 min) during MW drying were obtained. Clustering analysis among samples type and drying time categories was finally performed combining the selected features (shrinkage, brightness and saturation retentions). From results obtained by DIA analysis, several changes in shrinkage and colour features due to drying time and sample type were observed showing the potential of this methodology to be applied as an attractive and economical alternative to evaluate the appearance changes of foodstuffs during drying at industrial scale. K-means clustering analysis (k=3) showed poor results identifying drying time categories (adjusted rand index values up to 0.503). In the case of clustering by sample type (FR vs OD) dataset containing information only for long drying time category (>70 min) showed good agreements among true and k-means labels (adjusted rand index: 0.958, adjusted mutual index: 0.913).