Technological solutions for intelligent data processing in the food industry
The article is devoted to the possibilities of application of artificial neural networks (ANN), which are a mathematical model, as well as its software or hardware implementation, built on the principle of organization and functioning of nerve cell networks of a living organism. Convolutional neural networks are arranged like the visual cortex of the brain and have achieved great success in image recognition, they are able to concentrate on a small area and highlight important features in it. The widespread use of ANN in medicine for the evaluation of radiographs, blood pressure and body mass index of patients on the analysis of their retina is noted. The use of ANN in the food industry for input quality control of raw materials is promising. In the world practice, various methods of remote control of raw materials are used, for this purpose ultrasonic scanning devices are mainly used. Such devices and analysis systems control raw materials by the ratio of meat tissues (muscle, connective, fat) in the carcass or half-carcass, without affecting the tissue structure, do not lead the quality at the cellular (microstructural) level. It is established that the structure of muscle (diameter of muscle fibers, the safety of the cellular elements, the porosity of the tissue, integrity of muscle fibers) reflects the quality of the raw material, its thermal state. Our work has begun on the creation of an expert system for quality control of meat raw materials at the microstructural level using modern intelligent technologies as ANN and computer vision. This direction is relevant and socially significant in the development of the meat industry, as it will significantly speed up the process of analysis of the quality of raw meat in the research laboratories of meat processing enterprises and testing centers and improve the objectivity of the results.