The aim of the project is:
to complete the development of an already proven quality index method;
to incorporate the said method into hand held data terminals to be used: on board; for landings and sales; for stock control and production planning within the industries; and for fish inspection;
to implement a mechanized and rational computer vision based size sorting system, suitable for lean fish at arrival at the plant (information from this system regards length and weight of the fish, preferably combined with quality data concerning the actual raw material obtained by spot checks using the above mentioned quality index method);
and to establish a storage management system including a labelling system suitable for holding information from the said sorting system.
A study named 'Quality management of raw material in the food fish sector' has been dealing with the optimal use of the raw material. The final use of the raw fish in fish industries depends upon freshness of the fish and, in many instances, the size of the fish. An already proven quality index method has been adapted for cod, saithe, plaice, sardine and anchovies. Assessing schemes for each species have been made and incorporated into a computer program to be used in a hand held data terminal or in a personal computer (PC) for training. Test results from the schemes for cod and plaice are quite consistent, while results for saithe are less convincing. For anchovies and sardines the evaluation forms may have to be developed further, but the results look promising. Implementation of a computer vision length grading system in the filleting industry is a success. Length sorting by computer vision before filleting improved yield by 1 to 2%. Systems for labelling and traceability have been evaluated for the fish industry. Bar codes moulded into the fish box or placed in a slot are being used for identifying boxes while all information of the box's content is stored in a central database. A program to assist the plant manager in finding the best way of using the raw material has been made. It calculates the optimum contribution margin on basis of information on freshness, length of fish, weight, capacity, and product specification. The program is based on a model using a different principle from the first in first out system used in the industry today and looks superior to it. Difficulties have been encountered when transferring data to the hand held terminal chosen for the sensory assessments in this project, this not being 100% compatible with the standard PC.
Phase 1 is the development of programmes (software) for quality indexing of 4 or more important fish species based on storage experiments.

Phase 2 is the implementation of a visual size sorting system and a labelling system.

Phase 3 is the research and experiments for quality registration and indexing and labelling of the final products including the required information and a computer programme for management of the chilled storage holding the raw fish.