Research on Food Image Recognition of Deep Learning Algorithms

In recent years, food image recognition has been a research field that has received much attention. Diet is an essential part of people's daily life. Through food image recognition and classification, dietary evaluation is conducted to achieve the goal of healthy eating. Manual features and classification methods in traditional target detection algorithms can no longer meet the requirements of detection accuracy and speed, so it is necessary to introduce depth learning-based techniques into food image recognition. Deep learning is a complex machine learning algorithm that allows computers to simulate the human brain's thinking, judge new problems through data sets (human experience), and improve themselves to obtain relatively accurate models, just like humans. Among them, convolutional neural networks have developed most rapidly. They are applied to image recognition with the expectation of being able to automate recognition through machine learning, thereby reducing the workload of image recognition and providing more objective analysis results. Deep learning has achieved great success in intelligent image processing, image classification, and recognition and has excellent development potential and social value. This article implements automatic recognition of food images based on the recognition algorithm of the YOLOv5 convolutional neural network structure, which can detect food images with high detection efficiency. It explores the combination of deep learning to identify food images, providing a reference for people to eat healthily, especially for obese people to control calorie intake, and providing theoretical support for food image recognition.