SwimSense: Monitoring swimming motion using body sensor networks

Abstract

In order to effectively improve training quality of the swimmers, the activity monitoring technology based on body sensor networks (BSN) may be qualified for this task. In this paper, a monitoring system (SwimSense) for human swimming training locomotion based on BSN is established. SwimSense includes six measurement nodes, which can monitor the swimming strokes of several swimmers synchronously. The receiving node is connected with personal computer (PC) through USB cable, which allows the collected motion data can be transmitted to PC through wireless radio frequency communication, and the collected data can be used to motion analysis. The preliminary monitoring system mainly has two functions, at the first place, different swimming strokes may be recognized by using the monitoring system, and the selective classifier is Hidden Markov Model, and then according to the results of classification and the characters of different swimming strokes, phase segmentation of each swimming stroke is executed by using Support Vector Machine for the detailed research in the future.