Análisis digital de imágenes para evaluar el encogimiento de fresas sometidas a tecnologías emergentes de procesamiento

In fruits, strawberries in particular, appearance is an attribute of great importance. Appearance, which can be defined by color, shape, size and visual texture, not only determines the degree of maturation or the presence or absence of impurities, but also indicates the application of technological processes or prolonged storage conditions. In most cases, an excessive shrinkage of fruits negatively affects the preference and acceptance of these products by consumers. Particularly, in the case of dehydrated fruits, the geometric modifications (shape and size) caused by the process are a matter of great concern to the industry. Traditional techniques for monitoring physical and physico-chemical parameters of quality usually are time consuming, costly, laborious, invasive and impracticable during in-line monitoring at the industry. Computer vision techniques and digital image analysis are an attractive and economical alternative to evaluate the appearance changes of foodstuffs during processing. Therefore, the present work aims the evaluation and correlation of morphological characteristics area ( A ), by digital image analysis techniques, and thickness ( L ) measured by caliber in strawberries subjected to combined treatments that include application of edible coatings (alginate-lactate), osmotic dehydration (sucrose, 60 oB, 40 oC, 4h) and microwave assisted drying (1,2 W/g). Related to thickness, reductions of 60% (retention 40%) were obtained for fresh (FR+MW) and fresh coated (FR_R+MW) samples. OD pre-treated samples (DO+MW and DO_R+MW) reached values of 37% (retention 63%). The digital processing of images showed cross sectional area reductions up to 56% for FR+MW and FR_R+MW samples, being significatively higher (up to 18%) for samples previously osmo-dehydrated (DO+MW and DO_R+MW). By correlating the L and A percentage retention, during MW-drying, excellent correlations were attained for FR+MW and FR_R+MW samples (R 2 : 97 and 98%).