Development of Gesture Detection Application for the Exercise System and Accumulation of Reward Points

Abstract

The COVID-19 pandemic has affected the daily lives of global citizens. Currently, it was reported that COVID-19 would last longer. Therefore, living with the COVID-19 outbreak will be the new normal. The restriction of physical contact leads to less exercise, especially outdoor activity, but more activities online. Although it will reduce the spread of the disease, but it has a negative effect on people's physical and mental health, so our team has come up with an exercise system to collect points and redeem rewards by detecting gestures by using applications to increase the practice of physical activity and enhance the health of personnel in various organizations, even people in residence can have physical activity. There were two types of exercise postures to choose from: squat and bicep curl (arm muscle exercises), which can be divided into two levels: easy and difficult. We concluded that the application could calculate points correctly and was very accurate in counting reps, especially with the bicep curl, but there was still a slight discrepancy in the count of squat reps, which can be further developed to increase accuracy.