A method for assessing quality of food products based on optical spectra using intelligent pattern recognition in a full scale production environment

An Optical Fibre Sensor System has been developed for the online measurement of food quality both internally and externally as it cooks in a large-scale industrial oven. The quality of the food is determined by Its colour as it cooks and by automatically clarifying the food at different stages it is possible to determine whether the food is cooked to an optimum colour for profits. Development have been made on previous work by the authors by further examining the internal colour of the food and also looking at food that does not have cm oven colour externally. Spectroscopic techniques are employed to determine the colour and this signal is interrogated using an Artificial Neural Network. The resultant output spectral patterns can therefore be quantitatively Classified and thus categorized.