|Authors: ||R.P. Lima, S.M. Silva, R.L. Dantas, A.L. Dantas, A.S.B. Sousa, W.E. Pereira, R.M.N. Mendonça, G.H.C. Guimarães|
|Keywords: ||Ananas comosus, image segmentation, binarization, browning, ImageJ®|
The physiological disorder of translucency compromises the quality of fresh-cut pineapple.
Edible coatings have been applied to fresh-cut pineapple to reduce the translucency.
In general, a sensory panel using subjective scales performs the evaluation of translucency.
However, digital image processing, using free software such as ImageJ®, can enable more precise information to be obtained on translucency, considering, for instance, the tone of image coloration, the areas in which it occurs, and the pixels in the RGB or HSB channels.
Therefore, this study proposed a method for evaluation of translucency in fresh-cut 'Pérola' pineapple using digital image processing (DIP). 'Pérola' pineapple was fresh-cut into slices 10 mm thick that were coated with cassava starch at 3%, cassava starch 3% + fennel oil (0.025%), cassava starch 3% + 1% glycerol + 0.5% ascorbic acid, and cassava starch 3% + 1% glycerol + 0.5% ascorbic acid.
Coated slices were placed into cylindrical PET trays of 500 mL, which were covered with PVC film and stored for 6 days at 5°C and 90% RH. The translucency did not differ among slices treated with different coatings.
DIP allowed the segmentation and quantification of the translucent areas in fresh-cut 'Pérola' pineapple slices.
By the segmentation of the images, the HSB color system was more efficient in differentiating the translucent pulp than was the RGB system.
The translucency values obtained by DIP had a highly significant correlation (r=0.91**) with the values obtained by sensorial evaluations, as well as with the estimated values as a function of L* and Blue (r=0.93**).
Download Adobe Acrobat Reader (free software to read PDF files)