This is an outdated version published on 2022-01-12. Read the most recent version.
Preprint / Version 1

Validation of PITCHAI Markerless Motion Capture Using Gold Standard 3D Motion Capture


  • Tyler Dobos Department of Kinesiology, Faculty of Applied Health Sciences, Brock University, St. Catharines, ON, Canada
  • Ryan Bench Department of Kinesiology, Faculty of Applied Health Sciences, Brock University, St. Catharines, ON, Canada
  • Mike Holmes Brock University
  • Colin McKinnon 3MotionAI, Oakville, ON, Canada
  • Anthony Brady Driveline Baseball LLC, Kent, WA, USA
  • Kyle Boddy Driveline Baseball LLC, Kent, WA, USA
  • Michael Sonne 3MotionAI, Oakville, ON, Canada



Kinematics, Baseball, Pitching, Motion Capture


Most kinematic and kinetic assessments in baseball pitchers have been determined using marker-based motion capture systems. No current research exists on the feasibility of using single camera markerless motion capture technology for kinematic analysis of pitching. The purpose of this study was to compare and validate pitching kinematics (joint angles and summary metrics) from a markerless motion capture solution with a gold standard, 3D optical marker-based solution. 38 healthy pitchers of high school, college, or professional levels of experience threw 1-3 maximum effort pitches while concurrently using marker-based optical motion capture and pitchAI smartphone based (markerless) motion capture. Each pitch was time normalised from peak leg lift, to ball release. Time series pitchAI measures were compared to 3D motion capture using Pearson's R (R), R Squared (r2), and root mean square error (RMSE) for each joint angle. Discrete time points were evaluated for all joint kinematics at foot plant (FP), maximal shoulder external rotation (MER), and ball release (BR); as well as for descriptive metrics (stride length, arm speed, ball visibility, and ball path). For full time-series joint angles the pelvis and trunk had the best overall fit with an average r2 of 0.98, and 3.1 ± 1.1° of RMSE. The knee angles had an average r2 of 0.97 ± 0.02, and an average RMSE of 4.1 ± 2.0°. The throwing arm had an average r2 of 0.97 ± 0.02, with an average RMSE of 6.0 ± 2.2° across all measures.  The glove arm performed the worst, with an average r2 of 0.95, and 7.3° of RMSE. Across all discrete time points, the most accurate measures were the knees, followed by the trunk and pelvis, throwing arm, and finally glove arm. Stride length had an average RMSE of 3.90 ± 4.77%, and an r2 of 0.31. Arm speed had an average RMSE of 2.6 ± 3.4 m/s, and an r2 of 0.25. Ball visibility had an RMSE of 20.7 ± 24.1 ms, and an r2 of 0.10.  Ball path had an RMSE of 19.79 ± 23.9%, and an r2 of 0.45. When considering the technical ease of video-based solutions and an ability to measure in field, most metrics were within an acceptable range to the gold standard. pitchAI can be recommended as a markerless alternative to classic marker-based motion capture for baseball pitch kinematic analysis.


Metrics Loading ...


Barrentine, S. W., Matsuo, T., Escamilla, R. F., Fleisig, G. S., & Andrews, J. R. (1998). Kinematic analysis of the wrist and forearm during baseball pitching. Journal of Applied Biomechanics, 14(1), 24–39.

Birfer, R., Sonne, M. W., & Holmes, M. W. (2019). Manifestations of muscle fatigue in baseball pitchers: A systematic review. PeerJ, 7, e7390.

Birfer, R. (2019). “The Development of a Novel Pitching Assessment Tool. Brock University. MSc Thesis.

Boddy, K., Marsh, J., Caravan, A., Lindley, K., Scheffey, J., & O’Connell, M. (2019). Exploring wearable sensors as an alternative to marker-based motion capture in the pitching delivery. PeerJ, 7.

Brady, A., Briend, S., Caravan, A., Lindley, K., O'Connell, M. E., Gowdey, G., & Boddy, K. J. (2020, February 25). A Kinematic and Kinetic Comparison of Mound and Rocker Throws.

Cao, Z., Hidalgo Martinez, G., Simon, T., Wei, S. -E., & Sheikh, Y. A. (2019). OpenPose: RealTime Multi-Person 2D pose estimation using part affinity fields. IEEE transactions on pattern analysis and machine intelligence, 15, 1. 2019.2929257

Chen, C. H., and Ramanan, D. (2017). “3D human pose estimation= 2D pose estimation + matching,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (Honolulu, HI), 7035–7043. doi: 10.1109/CVPR.2017.610

Dillman, C. J., Fleisig, G. S., & Andrews, J. R. (1993). Biomechanics of Pitching With Emphasis Upon Shoulder Kinematics. Journal of Orthopaedic & Sports Physical Therapy, 18(2), 402–408.

Desmarais, Y., Mottet, D., Slangen, P., & Montesinos, P. (2021). A review of 3D human pose estimation algorithms for markerless motion capture. Computer Vision and Image Understanding, 103275.

Escamilla, R., Fleisig, G., Barrentine, S., Zheng, N., & Andrews, J. (1998). Kinematic Comparisons of Throwing Different Types of Baseball Pitches. Journal of Applied Biomechanics, 14, 1–23.

Fleisig, G. S., Andrews, J. R., Dillman, C. J., & Escamilla, R. F. (1995). Kinetics of baseball pitching with implications about injury mechanisms. The American journal of sports medicine, 23(2), 233-239.

Fleisig, G. S., Escamilla, R. F., Andrews, J. R., Matsuo, T., Satterwhite, Y., & Barrentine, S. W. (1996). Kinematic and Kinetic Comparison between Baseball Pitching and Football Passing. Journal of Applied Biomechanics, 12(2), 207–224.

Fleisig, G. S., Barrentine, S. W., Zheng, N., Escamilla, R. F., & Andrews, J. R. (1999). Kinematic and kinetic comparison of baseball pitching among various levels of development. Journal of biomechanics, 32(12), 1371-1375.

Fleisig, G., Chu, Y., Weber, A., & Andrews, J. (2009). Variability in baseball pitching biomechanics among various levels of competition. Sports Biomechanics, 8(1), 10–21.

Grantham WJ, Byram IR, Meadows MC, Ahmad CS. The Impact of Fatigue on the Kinematics of Collegiate Baseball Pitchers. Orthopaedic Journal of Sports Medicine. June 2014. doi:10.1177/2325967114537032

Gupta, V. (2019). Pose detection comparison: wrnchAI vs OpenPose. (April 15, 2020).

Ionescu, C., Papava, D., Olaru, V., & Sminchisescu, C. (2013). Human3. 6m: Large scale datasets and predictive methods for 3d human sensing in natural environments. IEEE transactions on pattern analysis and machine intelligence, 36(7), 1325-1339.

Kanko, R. M., Laende, E., Selbie, W. S., & Deluzio, K. J. (2021). Inter-session repeatability of markerless motion capture gait kinematics. Journal of Biomechanics, 121, 110422.

McKinnon, C. D., Sonne, M. W., & Keir, P. J. (2020). Assessment of Joint Angle and Reach Envelope Demands Using a Video-Based Physical Demands Description Tool. Human Factors, 0018720820951349.

Mündermann, L., Corazza, S. & Andriacchi, T.P. The evolution of methods for the capture of human movement leading to markerless motion capture for biomechanical applications. J NeuroEngineering Rehabil 3, 6 (2006).

Nakano, N., Sakura, T., Ueda, K., Omura, L., Kimura, A., Iino, Y., Fukashiro, S., & Yoshioka, S. (2020). Evaluation of 3D Markerless Motion Capture Accuracy Using OpenPose With Multiple Video Cameras. Frontiers in Sports and Active Living, 2.

Nicholls, R., Fleisig, G., Elliott, B., Lyman, S., & Osinski, E. (2003). Baseball: Accuracy of qualitative analysis for assessment of skilled baseball pitching technique. Sports Biomechanics, 2(2), 213–226.

Qiao, S., Wang, Y., & Li, J. (2017). Real-time human gesture grading based on OpenPose. Proceedings of the 10th International Congress in Image and Signal Processing, Biomedical Engineering and Informatics (CISP-BMEI).

Richards, J. G. (1999). The measurement of human motion: A comparison of commercially available systems. Human movement science, 18(5), 589-602.

Scarborough, D. M., Bassett, A. J., Mayer, L. W., & Berkson, E. M. (2020). Kinematic sequence patterns in the overhead baseball pitch. Sports biomechanics, 19(5), 569-586.

Seroyer, Shane T., et al. "The kinetic chain in overhand pitching: its potential role for performance enhancement and injury prevention." Sports health 2.2 (2010): 135-146.

Stodden, D. F., Fleisig, G. S., McLean, S. P., & Andrews, J. R. (2005). Relationship of Biomechanical Factors to Baseball Pitching Velocity: Within Pitcher Variation. Journal of Applied Biomechanics, 21(1), 44–56.

Tanaka, R., Takimoto, H., Yamasaki, T., & Higashi, A. (2018). Validity of time series kinematical data as measured by a markerless motion capture system on a flatland for gait assessment. Journal of biomechanics, 71, 281-285.

Thewlis, D., Bishop, C., Daniell, N., & Paul, G. (2013). Next-generation low-cost motion capture systems can provide comparable spatial accuracy to high-end systems. Journal of applied biomechanics, 29(1), 112-117.

Tian, Z. Z., Kyte, M. D., & Messer, C. J. (2002). Parallax error in video-image systems. Journal of Transportation Engineering, 128(3), 218-223.

Vafadar, S., Skalli, W., Bonnet-Lebrun, A., Khalifé, M., Renaudin, M., Hamza, A., & Gajny, L. (2021). A novel dataset and deep learning-based approach for marker-less motion capture during gait. Gait & Posture, 86, 70-76.

Veeger, H.E.J. (2000). The position of the rotation center of the glenohumeral joint. Journal of Biomechanics, 33, 1711-1715.

Werner, S. L., Fleisig, G. S., Dillman, C. J., & Andrews, J. R. (1993). Biomechanics of the Elbow During Baseball Pitching. Journal of Orthopaedic & Sports Physical Therapy, 17(6), 274–278.

Winter, D.A. (1990). Biomechanics and motor control of human movement. New York: Wiley Interscience.

Wu, G., Siegler, S., Allard, P., Kirtley, C., Leardini, A., Rosenbaum, D., Whittle, M., D'Lima, D. D., Cristofolini, L., Witte, H., Schmid, O., Stokes, I., ., & Standardization and Terminology Committee of the International Society of Biomechanics. (2002). ISB recommendation on definitions of joint coordinate system of various joints for the reporting of human joint motion--part I: ankle, hip, and spine. International Society of Biomechanics. Journal of Biomechanics, 35, 543–548.

Wu, G., van der Helm, F. C. T., Veeger, H. E. J. D., Makhsous, M., Van Roy, P., Anglin, C., Nagels, J., Karduna, A. R., McQuade, K., Wang, X., Werner, F. W., Buchholz, B., ., & International Society of Biomechanics. (2005). ISB recommendation on definitions of joint coordinate systems of various joints for the reporting of human joint motion--Part II: shoulder, elbow, wrist and hand. Journal of Biomechanics, 38, 981–992.

Yang J, Mann BJ, Guettler JH, et al. Risk-Prone Pitching Activities and Injuries in Youth Baseball: Findings From a National Sample. The American Journal of Sports Medicine. 2014;42(6):1456-1463. doi:10.1177/0363546514524699.