Modeling of fried potato chips color classification using image analysis and artificial neural network

Quality of potatoes in chips industry is estimated from the intensity of darkening during frying. This is measured by a human jury, subject to numerous factors of variation. Gray level intensities were obtained for the apex, the center, and the basal parts of each chip using image analysis of frying assays. We then tested a feed-forward artificial neural network designed to associate these data with color categories. It behaved in good agreement with human estimations, obtaining correlation coefficients of 0.972 for training data and of 0.899 for validation data. A systematic study of the response of the model allowed understanding the criteria of evaluation used by the human operators.