Effective training without risk of injury based on quantitative data is important for all athletes. In bicycle racing, pedaling skill analysis with electromyography (EMG) is gaining attention as a technique to use in addition to power training and heart rate training, which already have effective measuring devices. However, EMG is not easy for non-specialists to use because of the difficulties in analysis and interpretation. We, therefore, propose a multilateral pedaling skill evaluation technique to quantitatively and practically analyze with surface EMG wear for non-specialists. This technique can extract strong and weak pedaling skill items that are difficult to detect through visual observation. It also removes the need to acquire additional data only for preprocessing and is well adapted for measurement in an actual environment over a long period. We verify the technique's expressiveness and feasibility through experiments involving 14 participants. The results indicate its potential for pedaling skill training.
© Copyright 2018 ISWC '18 Proceedings of the 2018 ACM International Symposium on Wearable Computers. Alle Rechte vorbehalten.