A Nondestructive Method Based on an Artificial Vision for Beef Meat Quality Assesement

Abstract: The performance of artificial vision as a nondestructive technology has been evaluated in monitoring beef meat quality at a storage temperature of 4°C for more than two weeks. A reference method based on bacteriological measurement is performed in parallel with the artificial vision system to analyse the meat samples. Artificial vision data were collected from color image of meat samples in parallel with data from microbiological analysis for the enumeration of the population dynamics of total viable counts (TVC). Two color models are used to define fresh beef color in this study: the RGB (Red, Green and Blue) and HSI (Hue, Satutation and Intensity) model. Fuzzy ARTMAP artificial neural network based on a classification technique is used to investigate the performance of the artificial vision system in the quality classification of beef meat. The Fuzzy ARTMAP models built classified beef meat samples based on the total microbial population into “unspoiled” (microbial counts < 6 log10 cfu/g) and “spoiled” (microbial counts ≥ 6 log10 cfu/g). Good classification rates are obtained (95.24 %). Finally training and testing an artificial system will be considered as a useful alternative tool for beef meat quality assesement.

Keywords: Bacterial measurement; Artificial vision; Image analysis; Features extraction; Fuzzy ARTMAP classification model.