Research

Machine Vision II: Identifying Quality Indices in Salmon Fillets.

Problem: As the Alaska salmon industry produces more fillets as its primary value added product, additional quality grading for defects is essential. Objective grading of fillets for bruising and gaping using a machine vision system can assure consistent quality.

Approach: This project continues the development of machine vision for the Alaska seafood industry modifying the earlier system to evaluate and grade salmon fillets for bruising and gaping. Modification will include developing procedures for imaging these defects and adding analysis software to evaluate the images. The modified system will then be tested in a commercial operation. The objective of this project is to use our vision system in seafood plants for improving finished product quality grading for fillets.

Results: Improved imaging procedures were developed to better detect bruises and gaping in fresh and frozen salmon.  The optimum Angle of the fish to the camera, Illumination angle to the fish and the use of polarizing filters were evaluated.  A 45° fish angle, 90° lamp angle and polarized light were best at imaging gaping.  In addition, converting images to grayscale also enhanced the detection of gaping.  Bruising was more straight forward that the standard procedure developed in earlier project effectively images bruises.  Again the image conversion to grayscale enhanced bruise detection.

Presentations & Publications: 2009 Pacific Fisheries Technologists and Institute of Food Science Annual Meeting and peer-reviewed paper submitted to the Journal of Food Science.

Contact person: Dr. Chuck Crapo email: stsmiley@alaska.edu