Title: Research on intelligent recognition technology of gymnastics posture based on KNN fusion DTW algorithm based on sensor technology
Authors: Dong Fu
Addresses: Department of Physical Education, Shangluo University, Shangzhou, Shaanxi Province, China
Abstract: In order to realise the scientific and intelligent training of gymnasts, a research on the intelligent recognition algorithm of human movement and posture in the process of gymnastics using Kinect sensor technology is proposed this time. Firstly, Kinect sensor is used to obtain the basic data of human posture, and the relative distance and angle sequence of human joint points are taken as the basic characteristic parameters of posture recognition. After completing the training of the sample set, the KNN algorithm is used to recognise the gymnastics posture, match the standard target of the best angle curve and realise the evaluation of the movement. The DTW difference is used as the experimental parameter to obtain the final action score. This time, ten gymnasts are selected for simulation. The results of simulation analysis show that the proposed method can well realise the recognition and quantitative evaluation of gymnastic movements.
Keywords: gymnastics; body posture; catch; distinguish; Kinect sensor; evaluating.
DOI: 10.1504/IJWMC.2023.132423
International Journal of Wireless and Mobile Computing, 2023 Vol.25 No.1, pp.58 - 67
Received: 01 Jun 2022
Received in revised form: 29 Dec 2022
Accepted: 12 Jan 2023
Published online: 19 Jul 2023 *