Evaluation of 3D Markerless Motion Capture Accuracy Using OpenPose With Multiple Video Cameras

Abstract

There is a need within human movement sciences for a markerless motion capture system, which is easy to use and sufficiently accurate to evaluate motor performance. This study aims to develop a 3D markerless motion capture technique, using OpenPose with multiple synchronized video cameras, and examine its accuracy in comparison with optical marker-based motion capture. Participants performed three motor tasks (walking, countermovement jumping, and ball throwing), and these movements measured using both marker-based optical motion capture and OpenPose-based markerless motion capture. The differences in corresponding joint positions, estimated from the two different methods throughout the analysis, were presented as a mean absolute error (MAE). The results demonstrated that, qualitatively, 3D pose estimation using markerless motion capture could correctly reproduce the movements of participants. Quantitatively, of all the mean absolute errors calculated, approximately 47% were <20 mm, and 80% were <30 mm. However, 10% were >40 mm. The primary reason for mean absolute errors exceeding 40 mm was that OpenPose failed to track the participant's pose in 2D images owing to failures, such as recognition of an object as a human body segment or replacing one segment with another depending on the image of each frame. In conclusion, this study demonstrates that, if an algorithm that corrects all apparently wrong tracking can be incorporated into the system, OpenPose-based markerless motion capture can be used for human movement science with an accuracy of 30 mm or less.

Two-dimensional video-based analysis of human gait using pose estimation

“…Some prior studies have used OpenPose to investigate particular features of walking or other human movement patterns (Chambers et al, 2019;Sato et al, 2019;Viswakumar et al, 2019;Nakano et al, 2020;Ota et al, 2020;Zago et al, 2020). Our findings align with these reports in that we found OpenPose to be capable of reasonably accurate tracking of human movement (in our case, walking).…”

Marker-Less Monitoring Protocol to Analyze Biomechanical Joint Metrics During Pedaling

“…It is a marker-less system, and these systems currently still have some inherent limitations [14], [15]. It is estimated that the accuracy of the results given by OpenPose can be lower than 30 mm or less in case the algorithm corrects the tracking errors [31]. The maximum sample frequency of the webcams used in our experiments is 60 Hz, which could be a limitation when recording high speeds (> 120 rpm).…”

Computer Vision to Automatically Assess Infant Neuromotor Risk

“…However, before clinical interpretation of the 2-D movements into 3-D rehabilitation measures, there is a need for validation of pose estimation against 3-D sensor-based measurements. Past work has compared OpenPose estimates on adult movements to motion capture and reported mean absolute errors of less than 20 mm in ~50% data and less than 30 mm in 80% of data [47]. In the future, other studies could compare the pose estimates obtained from our pipeline on infant movements to 3-D motion capture to provide a better basis for clinical interpretability of the features obtained.…”