TABLE TENNIS ROBOT SYSTEM APPLYING DEEP LEARNING

摘要

According to embodiments of the present invention, the table tennis robot system to which deep learning is applied is capable of interactive rallying with the user using deep learning and a vision system, and the table tennis robot of the presented method detects both complex mathematical calculations and the trajectory of the ping-pong ball. It is differentiated from the previously proposed robots by being able to improve performance according to the user's exercise ability. In addition, it can be operated with just one stereo type camera, making it easy for users to access financially and deep learning. The table tennis robot used can make a big contribution to the rehabilitation problems of the elderly, patients and the disabled in the future.