artificial neural network
backpropagation learning
food processing industry
multilayer perceptrons
online color measurement
optical fiber sensors
premature browning detection
principal component analysis
self organizing maps
spectral classification
weighted distance fuzzy logic class membership
COLOR
CLASSIFICATION
QUALITY
SYSTEM
FOOD
This paper examines the design of an optical fiber sensor that monitors ground beef online, as it cooks, in order to determine the quality of the meat; in particular, if premature browning has occurred. The experimental work involved cooking fresh meat and meat that has been stored in a freezer for, one week, one month and three months, and recording the reflected spectra and temperature during the cooking process in order to develop a classifier, based on pattern recognition techniques that can determine premature browning and the degree to which the meat has been cooked. A comparison of this sensor is made with traditional research methods of detecting premature browning, to demonstrate that it would be more commercially viable as an online solution.