Food Quality Inspection
The food processing industry is constantly searching for automated methods in the area of quality monitoring and control. Firstly, production based on the quality of the raw materials will not only help optimise the value of the raw materials, but also will enable the provision of products with well-defined qualities for recipe optimisation e.g. defining the fat content of pork trimmings used for making different sausages. Secondly, documentation of product quality enables efficient streaming of products into different markets, e.g. fish that are higher in fat content obtain a higher price in Japan than in Europe. Furthermore, increased consumer awareness results in a higher demand for producers to provide documentation and traceability.
One important reason for the lack of such automated instrumentation is that food is a complex biological product to measure with variations in size, shape, texture and structure. Other challenges faced are the heterogeneity within the product and the increasing demand for pre-mortem measurements. It is also highly preferable in the food industry that all important quality parameters are measured contact free and while moving on a conveyor.
Together with Nofima we have solved such challenges for fish, crab, vegetables and meat, developing methods and instrumentation based on NIR transflection. Methods have been developed for measuring the water content of split cod, fat content of salmon and inhomogeneous meat trimmings and pieces, and water and fat content of fried potatoes. Preliminary investigations have also been made in measuring the ice fraction level of superchilled salmon and the core temperature of cooked and chilled small meat products, e.g. sausages.
Another area of interest is the development of low capacity, low cost systems for manual food quality inspection. SINTEFs DOE technology is well suited to this type of instrumentation and we expect that within few years both at-line and handheld instruments will be developed for applications such as fat, protein and water content of fish or meat, sugar content of fruit, quality parameters of wine, grain, dairy products, feed, etc.
Computer vision is a tool for objective assessment of visual qualities: colour, shape, defects, and other irregularities. We have made systems for precisely classifying the redness of salmon, locating pin bones in salmon fillets and have also developed methods for grading split cod by defining blood and liver stains, holes, shape irregularities. The advantage of replacing manual inspection by camera methods is not only automation, but also objective evaluation, reproducibility and traceability.
Research activities:
As ICT specialists we do most of our food related research in co-operation with Nofima (Fiskeriforskning and MATFORSK) and other SINTEF institutes.
After we have developed a prototype measurement system we cooperate with Qvision and other companies for industrialisation of the system.
If you are interested in more information please contact
Marion O'Farrell.
Fat and colour measurements of salmon by scanning transflectance.