Recognizing badminton hitting movements Based on Multi-sensor trajectory recognition algorithm
Li Li, Feng Li · 2023
In badminton, the swing motion exhibits variable acceleration of posture rotation, leading to an increase in the dynamic calculation error of the yaw angle. This, in turn, results in a greater cumulative error in subsequent swing trajectory recognition calculations. To address this issue, this paper proposes a multi-sensor fusion swing trajectory recognition method, denoted as TRM. The TRM method utilizes intelligent devices equipped with multiple sensors to collect sensor data during the swing process, enabling high-precision swing trajectory recognition. Experimental results demonstrate that the accuracy of the swing yaw angle and attitude angle is within 1 degree, while the average trajectory error is controlled within 10%. This method significantly improves the accuracy of badminton racket swing trajectory recognition, offering an effective solution for advancing badminton motion recognition technology.