Vision-based batting training system

Among all of the skills required in sports, batting a ball is one of the most difficult. In this paper, we propose a vision-based batting training system for batting practice. Our goal is to predict the flying trajectory of a batted ball by using a high-speed vision system and then display the best batting point in the batter's field of view immediately after the ball is pitched. We designed a trajectory predictor based on a Kalman filter algorithm, in which we assume respective equations of motion on each of the ball's axes. The proposed vision notification system is a combination of a Unity 3D scene and a wearable gadget.