HappyFit: a dietary calories counter application using image recognition

• Main Authors: Mohd Isnin, Nurulamira, Mohd Rum, Siti Nurulain, Mohd Ariffin, Noor Afiza, Asmawi, Aziah
• Format: Article
• Published: RMP Publications 2020
• Online Access: http://psasir.upm.edu.my/id/eprint/80032/; https://ijurjournal.weebly.com/volume-2-issue-1-2020.html

Over the last three decades, obesity has increased worldwide. Statistics have shown that the rate of obese population all over the world is keep increasing and the reasons are varied. This has led the attention of the World Health Organization (WHO) [1]. Studies have shown that peoples’ awareness of the risks being an obese are also improving. Today, people are becoming more concerned about their weight, and eat more healthy food and practicing a healthy lifestyle. Obesity treatment requires constant monitoring on daily calorie intake. With the evolution of information technology, the smart phone has become a vital gadget for everyday life. An Androidbased apps with calorie intake counter features can be very helpful to control obese. There are many mobile applications popping up like mushrooms that can assist people to count their calorie intake, however, most of them were specifically developed for western people and not suitable for Malaysian user. In addition, the applications do not provide any automated advisory for appropriate exercise based on the calorie intake of a user. This paper presents the development process of HappyFit, a mobile application that used the Convolutional Neural Network (CNN) for the food image recognition. This application is specifically developed for Malaysian users to monitor their calorie intakes based on the food images. This application is also able to provide the suitable exercise types based on the amount of the calorie intake by the user.