Automatic table tennis scoring system based on smart wearable devices

Abstract

Recently people's interests in health care are increasing, and many wearable device manufacturers have released various kind of fitness trackers for exercises and activities. However, most of them are focused on personal activities not on group activities. In this paper, we propose a system that automatically scores a table tennis game by using smartwatches and smartphones. The system analyses sensor data from the smartwatches, detects sound of a ball and calculates scores based on table tennis scoring algorithm. To evaluate the system, we collected 521 swing data and 50 ball sound data from 5 people, and the accuracy was 97.8%(510/521) and 82%(41/50), respectively. We expect that this system can be applied in a wide range of group activities with a low expense. © 2018.