Korean Journal of Applied Biomechanics (한국운동역학회지)

Korean Society of Applied Biomechanics (한국운동역학회)
권호 표지
DOI QR코드

DOI QR Code

Reliability and Consistency Analysis of Markerless Gait Motion using a Single Camera

단일 카메라를 활용한 마커리스 보행동작의 신뢰성 및 일치성 분석

  • Ji Su Yoon (Department of Physical Education, Graduate School of Pukyong National University) ;
  • Sang Hong Park (Department of Electronic Engineering, Pukyong National University) ;
  • Soo Ji Han (Industry-Academia Cooperation Foundation, Pukyong National University) ;
  • Sa Bin Chun (Department of Physical Education, Graduate School of Pukyong National University) ;
  • Beom Soo Kim (Department of Physical Education, Graduate School of Pukyong National University) ;
  • So Young Joo (Department of Physical Education, Graduate School of Pukyong National University) ;
  • Chang Hyeon Eom (Department of Physical Education, Graduate School of Pukyong National University) ;
  • Jong Chul Park (Department of Marine Sports, Pukyong National University)
  • Received : 2024.12.14
  • Accepted : 2024.12.19
  • Published : 2024.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

Objective: The purpose of this study is to evaluate the clinical usability of markerless motion analysis by comparing single camera-based markerless motion analysis using computer vision and infrared camera-based motion analysis. Method: This study was conducted on five healthy adults (age: 28.60 ± 5.24 years, height: 172.10 ± 7.06 cm, weight: 73.10 ± 14.07 kg). A single camera was installed in the sagittal plane, and the subject walked a total of 6 m along a path three times at his own pace. Position data of the joint centers of the shoulders, hips, knees, ankles, and toes and angle data of the lower extremity joints were calculated and analyzed. Results: Joint center position data showed high agreement between methods for anterior-posterior and vertical movements, excluding medial-lateral movements. However, the reliability of markerless-based motion analysis showed lower reliability toward distal joints, and consistent errors were confirmed in medial and lateral movements. It was confirmed that the average error in the angle data of the lower extremity joints increased toward the distal joints. Conclusion: Research results show that single camera-based markerless motion analysis using computer vision is effective for forward, backward, and vertical motion when photographed in the sagittal plane. However, large errors occur in depth data between the subject and the camera, which can affect other biomechanical variables. Therefore, it is effective in terms of real-time analysis, easy setup, and low cost when using a single camera, but it is emphasized that supplementation with depth data is necessary for accurate analysis.

Keywords

Acknowledgement

This project (result) is the result of a study conducted as part of the 3-stage Industry-Academia-Research Cooperation Leading University Development Project (LINC 3.0) supported by funds from the Ministry of Education and the National Research Foundation of Korea.

References

  1. Arrowsmith, C., Burns, D., Mak, T., Hardisty, M. & Whyne, C. (2022). Physiotherapy exercise classification with single-camera pose detection and machine learning. Sensors, 23(1), 363.
  2. Auer, S., Süß, F. & Dendorfer, S. (2024). Using markerless motion capture and musculoskeletal models: An evaluation of joint kinematics. Technology and Health Care, 32(5), 3433-3442. https://doi.org/10.3233/THC-240202
  3. Carse, B., Meadows, B., Bowers, R. & Rowe, P. (2013). Affordable clinical gait analysis: An assessment of the marker tracking accuracy of a new low-cost optical 3D motion analysis system. Physiotherapy, 99(4), 347-351. https://doi.org/10.1016/j.physio.2013.03.001
  4. Colyer, S. L., Evans, M., Cosker, D. P. & Salo, A. I. (2018). A review of the evolution of vision-based motion analysis and the integration of advanced computer vision methods towards developing a markerless system. Sports Medicine-Open, 4, 1-15. https://doi.org/10.1186/s40798-017-0116-x
  5. Hollman, J. H., McDade, E. M. & Petersen, R. C. (2011). Normative spatiotemporal gait parameters in older adults. Gait & Posture, 34(1), 111-118. https://doi.org/10.1016/j.gaitpost.2011.03.024
  6. Hulleck, A. A., Alshehhi, A., El Rich, M., Khan, R., Katmah, R., Mohseni, M., Arjmand, N. & Khalaf, K. (2024). BlazePose-Seq2Seq: Leveraging regular RGB cameras for robust gait assessment. IEEE Transactions on Neural Systems and Rehabilitation Engineering.
  7. Hulleck, A. A., Menoth Mohan, D., Abdallah, N., El Rich, M. & Khalaf, K. (2022). Present and future of gait assessment in clinical practice: Towards the application of novel trends and technologies. Frontiers in Medical Technology, 4, 901331.
  8. Kidziński, Ł, Yang, B., Hicks, J. L., Rajagopal, A., Delp, S. L. & Schwartz, M. H. (2020). Deep neural networks enable quantitative movement analysis using single-camera videos. Nature Communications, 11(1), 4054.
  9. Maynard, V., Bakheit, A., Oldham, J. & Freeman, J. (2003). Intra-rater and inter-rater reliability of gait measurements with CODA mpx30 motion analysis system. Gait & Posture, 17(1), 59-67. https://doi.org/10.1016/S0966-6362(02)00051-6
  10. McGinley, J. L., Baker, R., Wolfe, R. & Morris, M. E. (2009). The reliability of three-dimensional kinematic gait measurements: A systematic review. Gait & Posture, 29(3), 360-369. https://doi.org/10.1016/j.gaitpost.2008.09.003
  11. Moran, M. F., Rogler, I. C. & Wager, J. C. (2023). Inter-session repeatability of marker-less motion capture of treadmill running gait. Applied Sciences, 13(3), 1702.
  12. Moro, M., Marchesi, G., Hesse, F., Odone, F. & Casadio, M. (2022). Markerless vs. marker-based gait analysis: A proof of concept study. Sensors, 22(5), 2011.
  13. Ohri, A., Agrawal, S. & Chaudhary, G. S. (2021). On-device real-time pose estimation & correction. International Journal of Advances in Engineering and Management (IJAEM), 3(7), 1691-1696.
  14. Pirker, W. & Katzenschlager, R. (2017). Gait disorders in adults and the elderly: A clinical guide. Wiener Klinische Wochenschrift, 129(3), 81-95. https://doi.org/10.1007/s00508-016-1096-4
  15. Ripic, Z., Nienhuis, M., Signorile, J. F., Best, T. M., Jacobs, K. A. & Eltoukhy, M. (2023). A comparison of three-dimensional kinematics between markerless and marker-based motion capture in overground gait. Journal of Biomechanics, 159, 111793.
  16. Stenum, J., Cherry-Allen, K. M., Pyles, C. O., Reetzke, R. D., Vignos, M. F. & Roemmich, R. T. (2021). Applications of pose estimation in human health and performance across the lifespan. Sensors, 21(21), 7315.
  17. Uhlrich, S. D., Falisse, A., Kidziński, Ł, Muccini, J., Ko, M., Chaudhari, A. S., Hicks, J. L. & Delp, S. L. (2023). OpenCap: Human movement dynamics from smartphone videos. PLoS Computational Biology, 19(10), e1011462.
  18. Van Hooren, B., Pecasse, N., Meijer, K. & Essers, J. M. N. (2023). The accuracy of markerless motion capture combined with computer vision techniques for measuring running kinematics. Scandinavian Journal of Medicine & Science in Sports, 33(6), 966-978. https://doi.org/10.1111/sms.14319
  19. Wang, H., Su, B., Lu, L., Jung, S., Qing, L., Xie, Z. & Xu, X. (2024). Markerless gait analysis through a single camera and computer vision. Journal of Biomechanics, 165, 112027.
  20. Winter, D. A. (1991). Biomechanics and motor control of human gait: Normal, elderly and pathological.
  21. Yang, J. & Park, K. (2024). Improving gait analysis techniques with markerless pose estimation based on smartphone location. Bioengineering, 11(2), 141.
  22. Young, F., Mason, R., Morris, R., Stuart, S. & Godfrey, A. (2023). Internet-of-things-enabled markerless running gait assessment from a single smartphone camera. Sensors, 23(2), 696.
  23. Young, F., Wall, C., Graham, L., Stuart, S., Morris, R. & Godfrey, A. (2023). Lightweight markerless identification of temporal gait outcomes with BlazePose. Paper presented at the 2023 IEEE 19th International Conference on Body Sensor Networks (BSN), 1-4.
  24. Yoo, T. G., Choi, J. R. & Kim, D. W. (2012). Dynamics-based walking motion tracking using a single camera. January 2012 Journal of Electronic Engineering Volume 49 SC No. 1.
  25. Zhang, Y., Gan, J., Zhao, Z., Chen, J., Chen, X., Diao, Y. & Tu, S. (2023). A real-time fall detection model based on BlazePose and improved ST-GCN. Journal of Real-Time Image Processing, 20(6), 121.