Cross county sking synchroinized Video + sEMG data

Published: 04-03-2019| Version 1 | DOI: 10.17632/bwshbf2knp.1
Jan Sedlak


This database were used to evaluate algorithm for detection of movement cycles based on the video and sEMG signal. The movement cycle identification is necessary for the calculation of an average sEMG profiles which are used to analyze a muscle activity during various locomotion. This database contains 7 data sets every 20.5 second of synchronous sEMG and video signals recorded during cross-country skiing in natural environment. The database was stored in file. The MATLAB mat-file format was used to save data. Please cite as following: [1] Sedlák, J. - Špulák, D. - Čmejla, R. - Bačáková, R.B. - Chrástková, Martina - et al. : Segmentation of Surface EMG Signals. In Applied Electronics - 2013 International Conference on Applied Electronics. Pilsen: University of West Bohemia, 2013, p. 257-260. ISSN 1803-7232. ISBN 978-80-261-0166-6.


Steps to reproduce

Data set 1 - 7 deal with different style of cross-country skiing. The video data was recorded at frame rate 25 Hz with resolution 576x720x3. Surface EMG data, respective directly the linear envelope signal was captured. The analog full-wave rectification process and low-pass filtering (cut-off frequency 70 Hz) were used before the analog-to-digital signal conversion. The sampling rate was 200 Hz and resolution 8 bit. We measured 7 muscles: m. gluteus maximus dx., m. gluteus medius dx., m. vastus medialis dx., m. adductor longus dx., m. gastrocnemius caput medialis dx., m. tibialis anterior dx., m. gastrocnemius caput medialis sin. Proper points for placement of electrodes were identified by a physiotherapist. Detail description is reported in [1]. mat-file structure (data_x.mat): EMG_envelope - 4096 x 7 linear envelope of sEMG data EMG_muscle_name - 1x7 cell of muscle name EMG_fs - sample frequency of sEMG signal VIDEO - 512 frames of synchroneous video data VIDEO_framerate - frame rate of video recorder