Data acquired by a single object sensor for the detection and quality evaluation of table tennis forehand strokes

Sahar S Tabrizi; Saeid Pashazadeh; Vajiheh Javani

doi:10.1016/j.dib.2020.106504

Data acquired by a single object sensor for the detection and quality evaluation of table tennis forehand strokes

Data Brief. 2020 Nov 6:33:106504. doi: 10.1016/j.dib.2020.106504. eCollection 2020 Dec.

Authors

Sahar S Tabrizi¹, Saeid Pashazadeh², Vajiheh Javani³

Affiliations

¹ Department of Computer Engineering, Faculty of Electrical and Computer Engineering, University of Tabriz, Iran.
² Department of Information Technology, Faculty of Electrical and Computer Engineering, University of Tabriz, Iran.
³ Department of Sport Management, Faculty of Physical Education and Sports Sciences, University of Tabriz, Iran.

Abstract

Shadow-play, is an assistance solution for Table Tennis training, develops novice players' strokes and performing skills, and helps the players' brain to train in terms of the correct positioning and how the proper stroke technique feels. Most currently proposed training assistance systems are rarely used in actual applications, as they are expensive and their setup is complex. Thus, there is a need for a practical and low-cost intelligent system training assistance solution, as well as the possibility of using this solution comfortably to assist players. This paper specifies Forehand shadow play strokes movement and orientation sensory dataset for Table Tennis using a miniaturized low-powered, inexpensive and non- intrusive Inertial Measurement Unit (IMU) BNO055. We mounted the IMU on the center of a standard Table Tennis racket's surface. Eight novices, eight professional players, and three high ranked Table Tennis coaches participated in this research voluntarily. The Racket enabled us to collect players' strokes' time-series data responsively and sensitively. Collected sensory time-series data contains 1570 samples for the Basic, Topspin, and Push Forehand strokes of the players. Besides, all performed strokes were manually labeled and scored by the coaches simultaneously. The sensory dataset contains data from one 9-axis IMU (3- axis Accelerometer, 3- axis gyroscope, and 3- axis magnetometer) and Euler angles (roll, pitch, and yaw angles), mounted on the Racket. Based on the nature of the Forehand movements, the center of the surface was empirically determined to be the appropriate sensor placement in this experiment. We accomplished the collection of all samples under conditions that have been set by the coaches. The authors expect that the collected dataset can be used in a digital shadow-play coaching system to automatically send feedback to novice players when they practice shadow-play Table Tennis strokes individually.

Keywords: Classification; Inertial measurement unit; Object sensor; Regression; Sports activity analysis; Sports activity recognition; Table tennis forehand.