Abstract: As deep learning techniques are increasingly applied with greater depth and sophistication in the food industry, the realm of food image processing has progressively emerged as a central focus of research interest. This work provides an overview of key practices in food image processing techniques, detailing common processing tasks including classification, recognition, detection, segmentation, and image retrieval, as well as outlining metrics for evaluating task performance and thoroughly examining existing food image datasets, along with specialized food-related datasets. In terms of methodology, this work offers insight into the evolution of food image processing, tracing its development from traditional methods extracting low and intermediate-level features to advanced deep learning techniques for high-level feature extraction, along with some synergistic fusion of these approaches. It is believed that these methods will play a significant role in practical application scenarios such as self-checkout systems, dietary health management, intelligent food service, disease etiology tracing, chronic disease management, and food safety monitoring. However, due to the complex content and various types of distortions in food images, further improvements in related methods are needed to meet the requirements of practical applications in the future. It is believed that this study can help researchers to further understand the research in the field of food imaging and provide some contribution to the advancement of research in this field.