The lab standard, briefly
For a long time, the gold standard for measuring human movement lived in a dedicated motion lab. Patients wore dozens of reflective markers, surrounded by an array of infrared cameras that tracked each marker in three-dimensional space. The data is excellent, but the setup is expensive, time-consuming, and confined to a handful of research and specialty centers. Most clinicians have never had access to it.
What "markerless" actually means
Markerless motion capture removes the markers and the specialized camera rig. Instead, computer-vision models identify anatomical landmarks — joints, limbs, and key points on the body — directly from ordinary video, an approach made practical by deep-learning pose-estimation methods developed over the past decade. The software infers how the body is positioned and how it moves frame by frame, then derives the same kinds of metrics a lab would produce: joint angles, stride length, cadence, symmetry, and more.
Why it matters for clinics
The shift is less about the technology itself and more about access. When movement analysis depends only on a camera most people already carry, objective measurement can move out of the specialty lab and into everyday clinical settings — a physical therapy gym, a neurology clinic, or even a patient's living room. The barrier stops being equipment and starts being simply pressing record.
Understanding the trade-offs
Markerless methods are not magic, and being clear-eyed about their limits builds trust. Accuracy depends on the movement being measured, the camera angle, lighting, and how standardized the capture is. Published comparisons show strong agreement with laboratory systems for core gait measures. Against simultaneous marker-based capture, a multi-camera markerless system tracked lower-limb joint centers to within about 2.5 cm for all joints except the hip, which was 3.6 cm (Kanko et al., 2021), and open-source pose estimation from ordinary 2D video measured step timing and step length with small errors in healthy adults (Stenum et al., 2021). Reviews are equally clear about the limits: spatiotemporal measures generally look equivalent to marker-based capture, while joint-angle estimates call for more caution (Wade et al., 2022) — in Kanko's comparison, for instance, rotations about a limb segment's long axis agreed least well. We summarize the peer-reviewed evidence on our validation page. Used within those boundaries, markerless capture brings lab-grade thinking to places it could never reach before.
If you are comparing products, our buyer's guide to movement analysis apps lists markerless, sensor-based and lab options by category. For how a phone camera compares with wearables and force plates, see smartphone capture vs. wearables; for what validation a clinical tool should show, see what makes movement measurement clinical-grade.
References
- Kanko RM, Laende EK, Davis EM, Selbie WS, Deluzio KJ. Concurrent assessment of gait kinematics using marker-based and markerless motion capture. Journal of Biomechanics. 2021;127:110665. doi:10.1016/j.jbiomech.2021.110665
- Stenum J, Rossi C, Roemmich RT. Two-dimensional video-based analysis of human gait using pose estimation. PLOS Computational Biology. 2021;17(4):e1008935. doi:10.1371/journal.pcbi.1008935
- Wade L, Needham L, McGuigan P, Bilzon J. Applications and limitations of current markerless motion capture methods for clinical gait biomechanics. PeerJ. 2022;10:e12995. doi:10.7717/peerj.12995