Automated Obesity Detection and Classification via Live Camera Analysis

Abstract

Nearly 2 billion people in the world are fighting against "Corpulence" means in simple words we can say that they are facing a type of chronic diseases called "obesity or fatness." Not only the Obesity but also malnourished Underweight and Overweight are the today's concern. This problem is for all people including all genders and age. According to WHO the fat people in the world 2023 out of 2 billion people are : 650 million adults, 340 million adolescents, and 39 million children. Likewise, about 3-5% of adult population fall under a category of Underweight. Obesity causes the risk of hazardous diseases like Type 2 diabetes, Stroke, heart disease, etc. Underweight people can have health problems like Anemia, Osteoporosis, Feeling tired all the time, etc. Hence by considering all the situations we decided to help them out. We are using dataset which is collected from Kaggle contains the features like Age, Gender, Height, Weight, and BMI, etc. We are using input Dataset as well as Live Camera to become a user convenient. We are using algorithms like SVM, Random Forest, Decision Tree for dataset as input. Likewise, we are using Logistic Regression for live camera as input. All these algorithms will be used in training to check accuracy, precision, recall, f1_ score, etc. The algorithm which is more efficient will be used for testing.