An Automatic Calorie Estimation System of Food Images on a Smartphone

Abstract

In recent years, due to a rise in healthy thinking on eating, many people take care of their eating habits, and some people record daily diet regularly. To assist them, many mobile applications for recording everyday meals have been released so far. Some of them employ food image recognition which can estimate not only food names but also food calorie. However, most of such applications have some problems especially on their usability. Then, in this paper, we propose a novel single-image-based food calorie estimation system which runs on a smartphone as a standalone application without external recognition servers. The proposed system carries out food region segmentation, food region categorization, and calorie estimation automatically. By the experiments and the user study on the proposed system, the effectiveness of the proposed system was confirmed.