Food Quality Monitoring

A method is developed for the online measurement of the colour of food, on the surface and at the core, while cooking in a large-scale industrial conveyor belt oven. This method employs optical fibre technology. The colour is measured using normalised intensity of the reflected spectrum in the visible range of the electromagnetic spectrum between 440-750nm, when the food is illuminated with a tungsten halogen light source. Various classification techniques have been employed, including Multi-Layer Feed Forward Neural Network, Self-Organising Maps and k Nearest Neighbour.Principal Component Analysis is also employed as a method of feature extraction.