Abstract

With the rapid expansion of food trade, there is a growing concern about health issues. As a result, consumer demand for high quality food is increasing and food quality analysis has become a very important and interesting area of research. The application of biosensors in food analysis is promising because specific biosensors can be used to easily access the nutritional composition inside food products, including macronutrients, trace elements and other bioactive substances. Deep learning includes many different types of artificial neural networks, and convolutional neural networks can be used to extract external features of food products from images, such as shape, size, color, etc. Deep neural networks are able to generate predictive models using different food properties. In this paper, we aim to show how to combine biosensors and deep learning to assess food quality. We first focus on the process of generating predictive models by deep neural networks and the datasets required to train the models. Secondly, we focus on how to use convolutional neural networks to extract external features of food products and representative research work on biosensors for food nutrient content analysis. Finally, the paper summarizes and looks at the challenges and possible solutions of the approach in the field of food quality assessment.