Study at a glance
Summary
Disclaimer: The following summary details independent academic research. Theia3D is intended for research, sports performance and educational use only. It is not cleared as a medical device and is not intended to be used for the diagnosis, treatment, mitigation or prevention of any disease, injury or medical condition.
Editor's note: The paper does not report a specific Theia3D software version number.
Study Overview
Design
A concurrent, within-subjects comparison study evaluating agreement between marker-based and markerless (Theia3D) motion capture for sagittal-plane lower-limb joint angle, torque, and power during three jumping tasks in professional basketball players.
Participants
Twelve professional basketball players (age 19.93 years; height 1.84 m; mass 78.38 kg) recruited from university and professional teams, free of lower-extremity or waist trauma in the six months prior to testing and with at least six years of training experience.
Data Collection
Participants performed a standing vertical jump, a standing long jump, and a running vertical jump in a fixed sequence, with maximum effort on each attempt and at least three valid trials of each movement recorded. Marker-based data were captured using 10 Vicon MX-F40 cameras (100 Hz) with 28 reflective markers placed on the pelvis and lower limbs, synchronized with two embedded force-measuring platforms. Markerless data were captured simultaneously using six Teledyne FLIR Oryx10GigE cameras processed through Theia3D, with both systems' data streams time-aligned via a shared synchronization module.
Analysis
Hip, knee, and ankle joint angle, torque, and power in the sagittal plane were calculated for both systems using a consistent six-degrees-of-freedom biomechanical model in Visual3D, with data normalized to each participant's body weight and to 101 points across each movement cycle. Agreement between systems was assessed using Pearson correlation coefficients, root-mean-square difference (RMSD), maximum and minimum absolute errors, and statistical parametric mapping (SPM) to identify specific moments during each jump where significant differences occurred between the two systems.
Key Findings
- Knee and ankle joint angle, torque, and power showed very strong correlation between markerless and marker-based motion capture across all three jump types (correlation coefficients generally 0.92 or higher), with low measurement differences at both joints (joint angle RMSD under 8.2 degrees across all joints and movements, with ankle and knee notably lower).
- The hip joint showed the greatest variability of the three joints across all three measures tested, including joint angle (RMSD up to 8.2°), torque (RMSD up to 0.41 N·m/kg), and power (RMSD up to 1.76 W/kg), each considerably higher than the corresponding knee or ankle values.
- Statistical parametric mapping identified specific, significant differences between the two systems concentrated at or near maximum hip flexion in all three jump types, while knee and ankle joint angles showed significant differences only at isolated points during some of the movements and not consistently across all three tasks.
- Both systems showed a consistent pattern in which markerless motion capture estimated slightly greater joint flexion and extension than marker-based motion capture at the point of peak movement, a pattern the authors attribute in part to greater sensitivity of marker-based systems to skin and soft-tissue movement relative to the underlying bone during fast, high-effort jumping.
- The authors conclude that the markerless system demonstrates high overall validity for lower-limb jumping biomechanics, particularly at the knee and ankle, but that hip joint estimation and certain specific jump phases warrant further validation before the two systems are treated as fully interchangeable for these measures.
What This Means for Movement Science Researchers and Basketball Performance Teams
For researchers and performance staff working with basketball players, this study offers a specific, practically useful finding: markerless motion capture can be used with confidence for knee and ankle biomechanics during common jump-based performance assessments, two joints that are frequently central to evaluating lower-limb power output and landing mechanics in basketball. The hip joint finding provides an equally concrete, actionable caution: teams and researchers relying on hip joint angle, torque, or power from jump assessments, particularly around the deepest point of hip flexion, should treat those specific values with some added caution when comparing or combining data collected with marker-based and markerless systems, rather than assuming full interchangeability across every joint and movement phase tested.
The Role of Theia3D in Movement Analysis
Theia3D was the markerless motion-capture system directly evaluated against marker-based motion capture in this study, specifically for its accuracy and reliability in a sport-specific jumping context that had not previously been closely examined for basketball. The study's finding of strong agreement at the knee and ankle, paired with a specific, well-defined area for further validation at the hip, offers researchers and performance staff a practically useful, joint-specific picture of where markerless jump analysis for basketball currently stands on firm ground and where additional caution and future validation work are still warranted.
Study Limitations and Future Directions
The authors note that the hip joint showed the largest estimation error of the three joints studied, and they attribute this in part to the larger displacement of the hip joint center during movement, the greater thickness of skin and soft tissue around the pelvis, and the practical challenge of securely placing and tracking pelvic markers, all of which can affect the accuracy of hip kinetic calculations for both systems, not only the markerless one. They note that the two systems' differing underlying biomechanical models may also contribute to the differences observed, a source of discrepancy that is difficult to fully eliminate given how each system defines its joint centers and segments. The study was limited to a laboratory setting and to three specific jump types using a single Theia3D-camera configuration, so these findings may not generalize to other jump types, approach speeds, or playing surfaces encountered in actual basketball training or competition. The authors identify further validating hip joint kinetics specifically, ideally using a gold-standard reference method such as biplanar fluoroscopy, and expanding this kind of sport-specific validation work to a broader range of basketball movements, as important next steps.
Citation: Wei, L., Yang, C., Huang, X., Tu, L., Xu, Y., Li, M., & Hu, Z. (2025). Comparison of lower limb kinematics and kinetics estimation of basketball players during jumping with markerless and marker-based motion capture systems. Acta of Bioengineering and Biomechanics, 27(2), 133-145. https://doi.org/10.37190/abb/207092
Link: Read the full study
Does Markerless Capture Accurately Measure Basketball Jump Biomechanics?
This study of 12 professional basketball players finds that Theia3D markerless motion capture shows strong agreement with marker-based motion capture for knee and ankle joint kinematics and kinetics during jumping, with greater variability specifically at the hip joint. If your work involves basketball or jump-based sports performance research, talk to our team about applying Theia3D to your own studies.