Experimental Setup for Markerless Motion Capture and Landmarks Detection using OpenPose During Dynamic Gait Index Measurement

The majority of motion analysis tools used in sports biomechanics and rehabilitation do not provide autonomous kinematic data collection without the use of markers, experimental settings, or extensive processing durations. These limits can make it difficult to employ motion capture in routine traini...

Full description

Saved in:
Bibliographic Details
Published inInternational Conference on Space Science and Communication (Print) pp. 286 - 289
Main Authors Abd Shattar, Normurniyati, Gan, Kok Beng, Abd Aziz, Noor Syazwana
Format Conference Proceeding
LanguageEnglish
Published IEEE 23.11.2021
Subjects
Cameras
Dynamic Gait Index
Dynamics
Face recognition
Head
Magnetic heads
Markerless motion capture
OpenPose
Tracking
Training
Online AccessGet full text
ISSN2165-431X
DOI10.1109/IconSpace53224.2021.9768699

Cover

More Information
Summary:The majority of motion analysis tools used in sports biomechanics and rehabilitation do not provide autonomous kinematic data collection without the use of markers, experimental settings, or extensive processing durations. These limits can make it difficult to employ motion capture in routine training or rehabilitation situations and there is an evident need for the development of automated markerless systems. It is vital to select the right technical equipment and algorithms for accurate markerless motion capture. This paper discusses the markerless human motion capture method for landmark detection using the open-source library OpenPose during the Modified Dynamic Gait Index (m-DGI) assessment. The m-DGI research covered the following phases: gait level surface, change in gait speed, gait with horizontal head turns and gait with vertical head turns. The four stages in m-DGI were captured using two Canon EOS 90D DSLR cameras. The videos were recorded in frontal and side views (subjects moved toward the front camera) and the source code of OpenPose's human skeleton identification system can extract the image into a sequence of frames from the recorded video. This method processes every frame of extracted images in OpenCV-Python to detect the human skeleton key points. The application of this methodology in research and clinical practice has the potential to provide easy, time-efficient, and perhaps more relevant assessments of human mobility.
ISSN:2165-431X
DOI:10.1109/IconSpace53224.2021.9768699