Implementation of Human Pose Estimation Using Angle Calculation Logic on The Elder of The Hands as A Fitness Repetition

Abstract

Sport cannot be separated from everyday life. Sport is a hobby for every age group. Everyone needs exercise to keep the body healthy and fit. Many people are busy with their respective activities, so they need more time to exercise and go to the gym. Apart from having to adjust the time, People stuck at Home do not have free access to go the Gym during the global pandemic. This can be overcome by tracking people's movements virtually. Human Pose Estimation can track the movements being carried out by a person in time. Data collection in the form of optimal detection distance, Human Pose Estimation in calculating the number of fitness repetitions, and data from the application of Human Pose Estimation when someone does fitness sports. The optimal distance webcam detects Human Pose Estimation for fitness reps is three meters. Webcams successfully track and predict the movements made by a person. The program can calculate fitness reps based on the angle at the elbow. The key points on the human body are connected by utility lines to form the body’s skeleton. Application, The repetitions in this detection utilize the value of the elbow angle when the hand is straight and bent. If the tip is >300, such as lifting dumbbells, then a stage up is counted as one rep. Meanwhile, when the angle forms > 1700, a stage down and one repetition also count The use of Media pipe for detection results is accurate and effective.