Image characterization of potato chip appearance during frying

The aim of this study was to evaluate the suitability of a computerized image analysis technique (with a flatbed scanner for image acquisition) in order to measure the amount and distribution of the most important visual aspects of potato chips: colour components (L*, a* and b*) and brown and oily areas on the surface. The potato slices were fried at three temperatures (170, 180 and 190 C) and times (2, 3 and 4 min). Pre-processing, segmentation and colour analysis were carried out by a software programmed in Matlab v6.5. Results showed a high linear relationship (R2 > 0.962) between image RGB values and those measured by conventional colorimeter. The applied image analysis technique was able to differentiate with high sensitivity among potato chip colours after the frying processes. On average the percentages of normal, brown and oily areas detected on the samples were 53.24, 24.04 and 22.96%, respectively. While data of brown area appearance were congruous with browning incidence bound to the entity of the frying process, discrepant results were obtained using the objective colour pattern for the evaluation of the extension of oily area in the surface of potato chips. Anyway, this technique presents a high potential to develop a computer vision on-line system for frying process optimization, as a function of the fat content of the final product.