Authors: Li Cao; Chongjiang Zhan
Addresses: Department of P.E. and Art Education, Zhejiang Yuexiu University of Foreign Languages, Zhejiang, Shaoxing, China ' Jiyang College of Zhejiang A&F University, Zhejiang, Shaoxing, China
Abstract: The purpose is to better mine the fitness motion data for intelligent wearable devices and promote the development of the new community fitness mode. First, the defects of the traditional fitness motion recognition system are analysed. Then, software engineering technology and Deep Learning (DL) technology are used to build a multi-layer fitness motion monitoring system. Finally, the data of running, riding, race walking, and rope skipping in the PAMAP2 data set are used for system evaluation. The results show that the proposed motion data monitoring system has an average accuracy of 97.622%, an average precision of 96.322% and a recall rate of 96.021% for fitness data recognition. The experimental results suggest that intelligent wearable devices with the proposed monitoring system can effectively mine wears' motion data and promote the development of the new community fitness mode.
Keywords: AIoT; motion recognition; intelligent life technology; intelligent wearable device.
International Journal of Grid and Utility Computing, 2022 Vol.13 No.1, pp.57 - 65
Received: 01 Feb 2021
Accepted: 22 May 2021
Published online: 11 Mar 2022 *