Color determination in dehydrated fruits: image análisis and photocolorimetry

MARCELA AGUDELO; NURIA ACEVEDO; CAROLINA SCHEBOR; MARÍA DEL PILAR BUERA

Mar del Plata

Congreso; Interim Meeting of the International Color Association AIC 2010 “Color and Food”; 2010

International Color Association

Color is a property that determines food quality and consumer acceptance. Dehydrated fruits are considered to be highly stable. However, they are prone to suffer discoloration during storage. Many natural pigments are unstable in dried media, and also brown pigments can be formed. These color changes cause deleterious changes in food appearance and organoleptic quality, and may be an indication of the decreased nutritional and functional properties of foods. Food discoloration can occur homogeneously, but most of the times heterogeneous distribution of color is observed. The objective of this work was to evaluate the kinetics of colour changes in different dehydrated fruits as a function of relative humidity, especially considering heterogeneous discoloration. Color changes of freeze-dryed apple, pear, melon and strawberry slices were determined using a computer vision system and a photocolorimeter. Color images were taken in a standardized grey box using a D65 illuminant, and converted to the CIELAB coordinates L*, a*, b*. The photocolorimeter measurements were performed at 2º observer and D65 illuminant. Dehydrated fruits surface were stored at several relative humidities at 45º C and color evaluation was performed at selected times. Apple, pear and melon systems presented homogeneous color distribution. Due to browning during storage, a* and b* values increased in these systems and L* values decreased. Strawberry samples, on the other hand, presented an heterogeneous color distribution. Therefore, for the image analysis the slices were divided in 3 different zones according to their L* values and were classified as dark, intermediate and light areas. L* values decreased after humidification at 20ºC and during storage at 45º C. In these samples the heat treatment caused anthocyanin pigment destruction, therefore a* values decreased in the dark and intermediate zones. The lighter areas did not show important changes in the a* values, being b* the less informative variable. The pixels proportion of light areas in the pictures was greater at low RH and short storage times. As increasing RH and storage time the pixels proportion of dark areas and the color distribution homogeneity increased. Photocolorimetry is a simple and easy technique to evaluate color changes, but it shows a limitation for the analysis of small areas or of heterogeneous samples. Computer vision analysis is a useful tool to assess optical properties of fruits based on its simplicity, and also allows the analysis of heterogeneous materials.