An Approach to Estimate Body Mass Index Using Facial Features

Abstract

People must have a healthy daily routine in order to stay away from health-related problems. They should keep their body measures in check. A simple measure like body mass index (BMI) can be considered as an indicator for the ratio of weight and height. Generally, health risk is directly proportional to the BMI value, so it is advised that one should keep check on it. But, due to the busy routine, normally people are reluctant to measure weight and height regularly and then observe the BMI value. This motivated us to come up with a system that aims to calculate the BMI by just using a face image of an individual. From the face image, the facial features are extracted using three methods. These are then used to estimate the BMI of the person using machine learning algorithms. In regression, seven algorithms and in classification, nine algorithms were tested. The multiple linear regression technique outperformed the other methods. The main objective of this study is to estimate the BMI from face image accurately in a hassle-free manner.