At major fish markets many landings are processed using automated grading machines where every fish is automatically weighed and recorded, providing Grading Machine Fish Weight (GMFW) data. However, to ensure the data can be efficiently and confidently used for fisheries advice, an effective system for data exchange still needs to be developed, and some data issues need to be resolved. These relate to marketing practices where some species are processed together; machine error where ice and/or individual fish stick together and are mistaken for individual fish; and errors if part of a landing is mislabelled. If grader machine technology were combined with automated image capture, the size of the fish could be calculated from the image and compared with the weight to provide biological parameters for fisheries science and condition factors for marketing. Species identification using Machine Learning (ML) would automatically resolve data issues and quality assure the data. If these data were made available for scientific purposes it could improve efficiencies and the spatial and temporal resolution of the biological data currently being collected. These images will also help develop machine learning and camera systems that can be used at sea.
The public and consequentially buyers are increasingly interested in the provenance of the fish they purchase. Assurances about the sustainability of fishing practices and fisheries management and the back story about capture to plate have become important features of marketing fish products. Information about fish quality and related pricing is potentially useful to the catching sector as well as through the value chain. Cefas aim to work alongside Plymouth Trawler Agents (PTA) to investigate the development of this camera interface and how the data can contribute to an integrated information system that brings commercial benefits to the industry.
Centre for Environment, Fisheries and Aquaculture Science (Cefas)