Undesired variations in the quality of dough for bread-making, resulting from variations in the process or recipe, may cause significant changes in the characteristics of the final product. However, these changes may be compensated by appropriately modifying the bread-making conditions: this requires the availability of particularly efficient process control equipment. In this context, a recent survey made by a group of German researchers (Paquet-Durand et al., 2012), is proposed a system based on the use of image analysis techniques. In particular, the test was carried out using a camera capable of continuously detecting the images inside the oven (at 3 seconds intervals), during product cooking. The samples were cooked at 230°C for 15 minutes in an electric oven whose original lights were replaced with LED lights. The authors demonstrate that to monitor the product’s changes in shape during the process, it is necessary to submit the images to rather complex algorithms. In the specific case a pre-existing tool has been changed, the so-called Viola-Jones algorithm. This tool has proven to be able to differentiate dough for bread from that for other bakery products (croissants) with an error of 5.6%. Assuming that the product is characterized by an elliptical form, the system easily anticipates the volume alterations, causing changes in dough’s height and width. In the final stage, the color of the product is tested by examining the images through RGB analysis based on artificial neural network. In conclusion, the authors affirm that through this developed tool it is possible to change cooking parameters in real time, while maintaining constant quality of the final product, and also energy consumption of the process.
O. Paquet-Durand et al., Journal of Food Engineering, 111, 2012, 425-431