Color evaluation of the gills and eyes of fish: machine vision and image analysis methods

Creative Commons License

ALÇİÇEK Z., Balaban M. O.

Institute of Food Technologists (IFT) Annual Meeting & Food Expo, Las Vegas, United States Of America, 25 - 28 June 2012, pp.252

  • Publication Type: Conference Paper / Full Text
  • City: Las Vegas
  • Country: United States Of America
  • Page Numbers: pp.252
  • Çanakkale Onsekiz Mart University Affiliated: No


1. Justification
The color of the gills is an indication of the “freshness” of fish, and is used subjectively in the
Quality Index Method (QIM). Similarly, the color of the eyes and their transparency /
opaqueness is an indication of fish freshness and used in QIM. The automation and objective
evaluation of these criteria is desirable.
2. Objectives
Our objective was to develop region-of-interest (ROI) methods to evaluate the color of the gills
region and eyes of the fish during storage on ice.
3. Methods
Snapper (Chrysophrys auratus) was obtained while fresh from a local market, and stored on ice for one week. The gill plates of the fish were cut out, exposing the gills. Every day, fish were
removed from the refrigerator, and placed in a light box with polarized lighting. A Nikon D300
digital camera, with a circular polarizing filter was used to obtain the images. The importance of using polarized lighting to obtain accurate color information will be illustrated. Software was
developed to allow either circular or polygonal ROI definition on the images. These ROIs could
be saved and later used again to always analyze the same location on the fish. Circular ROIs
were used for the analysis of the eyes, and polygonal ROIs for the gills.
4. Results
The average a* values of the gills were reduced from 32 to 24 during the first day, and to 17 after day 7. The average L* values of the eyes started at 20, and increased steadily to 34 throughout storage. Therefore this method can be used to quantify the color change of the eyes and gills of the fish during storage.
5. Significance
The visual evaluation of the QIM method can be achieved by machine vision / image analysis.