The best window selection of electromyography signal during riding motorcycle using spectrogram.

• Main Authors: Tengku Zawawi, Tengku Nor Shuhada, Abdullah, Abdul Rahim, Mohd. Saad, Norhashimah, Sudirman, Rubita, Rashid, Helmi
• Format: Article
• Language: English
• Published: Ismail Saritas 2023
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://eprints.utm.my/105736/1/RubitaSudirman2023_TheBestWindowSelectionofElectromyographySignal.pdf; http://eprints.utm.my/105736/; https://ijisae.org/index.php/IJISAE/article/view/3206/1793
• Institution: Universiti Teknologi Malaysia

Electromyography (EMG) signals are widely used as an important tool which helps to understand human activities. However, EMG signal has the complexity of random signals, highly nonlinear, non-stationary, and multi-frequency properties. Previous researchers have applied the time domain and frequency domain, but it lacks either time or frequency information, thus time-frequency distribution (TFD) such as Spectrogram is suitable and widely used in extracting EMG signals. However, this method using Hanning Window is a fixed window that compromises between time and frequency resolution. Some researchers used time window selection in their research, however, there are no standard guidelines for determining window selection for all EMG signals. Thus, this paper has presented the guidelines for determining the best window size for EMG signal while riding a motorcycle using Spectrogram. There are eight muscles for left and right from four types of muscles group which are Extensor Carpi Radialis, Trapezius, Erector Spinae, and Latissimus Dorsi. Six window sizes of 128, 256, 512, 1024, 2048 and 4096 ms are selected to determine the best size window to be used for the future analysis of the EMG signal. Machine Learning of SVM is used for confirmation performance evaluation for the best window size as the highest accuracy results. The results have proved window size 1024 is the best window size for EMG signal for riding a motorcycle. From this finding, the future analysis of this signal will use this size window when involving Spectrogram method.in the future.