| （英） |
his paper discusses a sensing system that can estimate different modes of bicycle manipulation during a bike ride. The system considers the following six modes of bicycle operations: (1) stopped and off-bike, (2) stopped and on-bike, (3) pedaling on bike, (4) riding without pedaling, (5) stand-up riding, and (6) braking. In our field trial, a bike rider wore an acceleration sensor at the frontal side of his waist so that the system can capture acceleration data during different kinds of bike rides and estimate bike-operation modes using Hidden Markov Model. As a result, classification was attained to some extent. |