3D (three-dimensional) analysis measures movement in all three spatial directions rather than within a single image plane. It can describe rotation, side-to-side movement, and movement toward or away from a camera that a single 2D view cannot fully capture.
Common 3D Approaches
- Marker-based optical motion capture: Several calibrated cameras, often infrared, track reflective markers placed on anatomical landmarks
- Multi-camera video: Two or more synchronized, calibrated cameras record the same movement, and landmark positions are reconstructed in 3D
- Markerless systems: Pose estimation is applied to multiple camera views to estimate 3D joint positions without markers
- Inertial sensors: Wearable units containing accelerometers and gyroscopes estimate segment orientation
- Depth sensing: Sensors that measure distance to surfaces, usually over a short range
What 3D Analysis Can Provide
- Joint angles in the sagittal, frontal, and transverse planes at the same time
- Segment rotations, such as pelvis and trunk rotation in a swing
- Positions and velocities that are less affected by camera angle
- When combined with force plates, inputs for estimating joint moments
Limitations
3D systems require careful calibration, controlled capture volumes, and consistent marker or sensor placement. Markers sit on the skin, which moves relative to the underlying bone, and modeling choices affect the reported joint angles. Laboratory systems can be expensive and may change how naturally an athlete moves. Results from different systems or models are not always directly comparable.
3D Versus 2D
2D analysis with a single phone or tablet is far more accessible and is often sufficient when a movement occurs mainly in one plane and the camera is placed perpendicular to it. Recording separate front and side views adds useful information, but two unsynchronized, uncalibrated views are not the same as a 3D reconstruction. Single-camera tools that estimate 3D pose from ordinary video infer depth rather than measure it, so their output should be checked carefully.
KineVision works with 2D video and 2D skeleton tracking. For many coaching questions, a well-positioned 2D recording is the practical choice, while 3D analysis is most valuable for research, detailed rotational analysis, and cases that require measurement across several planes.