Abstract:
Analysis of motorcycle rider‘s posture using sEMG signals
ABSTRACT
Comfort on a motorcycle is an important factor for people who ride regularly, especially for those who
travel over long distances. The riders are subjected to prolonged discomfort in upper body parts, which can
easily be prevented using posture correction. Other problems like poor road conditions, improper helmet
selection and its continuous use lead to pain in neck, shoulder and lower back. This work attempts to study
the fatigue developed in muscles such as Latissimus Dorsi (LD) and Flexor Carpi Ulnaris (FCU) during
motorcycle riding using Surface Electromyography (sEMG) signal. The study was conducted on 20 healthy
male subjects in a pre-determined track of 3kms. Number of rounds was fixed after pilot study. The signals
are recorded after each round. The signal features namely root mean square (RMS), mean frequency and
entropy is extracted from the recorded signals. Support Vector Machine (SVM) is trained using these signals
and the signals are classified as fatigue or non-fatigue. The accuracy of this classifier is found to be
74.26%.Precision is 0.7374 , recall and F1 score are 0.7426 and 0.7393. Area under curve (AUC) is
determined by taking average of each plot and is found out as 0.8529. The results show that majority of the
participants have significant muscle fatigue in LD after 3 rounds. An increase of 18.7% in RMS, 5%
decrease in mean frequency and a 6.04% decrease in entropy is found. Significant changes in these features
are observed after each round. This study can be used for posture correction analysis in future.
Keywords: Surface electromyography, muscle fatigue, motorcycle
