Host: The Japan Society of Mechanical Engineers
Name : [in Japanese]
Date : June 01, 2022 - June 04, 2022
In recent years, with the development of MEMS, IoT, machine learning and robot technology, these technologies have been used for coaching team sports. These coaching systems can be classified under two categories: one is to analyze the performance after the game, and the other is to present additional information during the game. These systems for team sports have not been studied for the purpose of learning an appropriate formation for the whole team by moving each player during the game. In this paper, We conducted two experiments to validate the effectiveness of the proposed image processing system for a team sport coaching system: 1) an experiment to verify the accuracy of detecting the position of a stationary person in real space, 2) an experiment to verify the recognition accuracy of four color-tagged human models moving in virtual space.
