49B-5 |
|
A. C. M. OLIVEIRA, Fishery Industrial Technology Center, Univ. of Alaska, Fairbanks, School of Fisheries & Ocean Sciences, 118 Trident Way, Kodiak, AK 99615-7401, M. O. Balaban, Dept. of Food Science & Human Nutrition, Univ. of Florida, Food Engineering Pilot Plant, PO Box 110370, Gainesville, FL 32611-0370, K. M. Portier, Dept. of Statistics, Univ. of Florida, 522 McCarty Hall C, PO Box 110339, Gainesville, FL 32611-0339, and S. F. O'Keefe, Dept. of Food Science & Technology, Virginia Polytechnic Institute & State Univ., 119 FST Bldg., Mail Code 0418, Blacksburg, VA 24061-0418. Machine vision is a new method of measuring color. Its comparison with traditional hand-held color meter was performed using Gulf sturgeon fillets. Gulf sturgeons were fed 3 artificial diets (Silver Cup Trout Pellets, Hybrid Bass Pellets and Catfish Fish Food Pellets). Fifteen sturgeons (5 fish/diet) were killed, eviscerated and stored in ice water for 3 days to undergo rigor. Fillets were placed in individual Ziploc bags, and stored on ice. Visual differences in color were observed for sturgeon fed trout diet, which imparted a yellowish-orange coloration to the flesh. A color study was performed to quantify these differences, and monitor color changes during ice storage. L*a*b* values were measured at days 0, 5, 10 and 15 using CR-200 Chroma Meter and machine vision. Their results were compared using a statistical mixed linear model with random effects for fish within treatment. Mean comparisons for L*a*b* values performed out using Least Squares Difference test with Bonferroni’s adjustment. Effects of treatment from L* and b* values were the same for both instruments. For a* values CR-200 showed no significant difference between treatments while machine vision detected significantly higher values for fish fed trout diet. Also, for CR-200 there was a clear change in L*a*b* values from day 0 to day 5 for all treatments, afterwards changes were not significant. L* values of trout treatment from machine vision increased significantly at every interval, while for catfish treatment the only significant change happened between day 0 and day 5. Trout treatment showed more fading than the other two treatments. Machine vision results also showed that a* and b* values did not vary within treatments considerably over 15 days of storage. Machine vision can detect color changes in non-uniform colored foods that the traditional color meters may miss.
Session 49B, Aquatic Food Products: General
|