Markerless Motion Capture

Simi Shape 3D allows full-body 3D motion capture (MoCap) and analysis without any markers! Simi Shape captures movement based on silhouettes.

What makes Simi Shape 3D unique?

The possibility to work markerless is advantageous in various ways:

  • no time loss due to attaching markers
  • less failure potential due to misplacement of markers
  • no artefacts because of markers moving on skin or clothes
  • no disturbance of movement caused by markers
  • hybrid tracking possibility
  • automatic tracking and processing

With Simi Shape you can use all possibilities of Simi Motion regarding data acquisition (synchronised high-speed video, kinematics, force plates, EMG, pressure, etc.), data processing, and data analysis.

Simi Shape integrated with Simi Motion offers the unique solution for hybrid use of two technologies, marker-based and markerless. While capturing markerless it can use additional markers, for example for certain rotation axes or for validation of results.

What is the working process with Simi Shape 3D?
  • Person is being captured by 8 cameras.
  • Silhouettes are extracted (separation of person and background).
  • Fitting of silhouettes for virtual 3D-model and person.
  • Optimal fitting of silhouettes with 3D-model allows extraction of 3D joint positions and joint angles.
  • The results are 3D joint positions and joint angles (similar to marker-based systems but without using markers)
Setup and system requirements:
  • Simi Motion System with 8 cameras
  • Cameras must be positioned around the subject from different angles/perspectives
  • Background seen by cameras must be stable and have good contrast to subject
What results do I get with Simi Shape 3D?

After automatic tracking with Simi Shape 3D you get 3D joint coordinates and rotations of all major body segments. This data is similar to the data being calculated by marker-based models, just without using markers.

Unique applications:
  • Record 3D movement of virtually any sport without interaction with the subject(s)
  • Capture 3D data of multiple subjects within the same recording
  • Recording can be done indoor and outdoor in a variety of environments, from indoor gymnastics to ski jump at -10 degrees
  • Constantly increasing library of automated protocols for movement in tennis, football, baseball and more…
  • Swim start (or similar) biomechanics can be obtained with the ability to record above and below the water
How accurate is Simi’s markerless system?

Simi Shape, is the world’s first and only 3D markerless motion capture system with “gold-standard” accuracies. Based on multiple high-speed cameras mounted around the capture volume, Simi’s algorithms are able to capture a biomechanical model, based on 21 Segments and 6 degrees of freedom. Simi’s cameras calibrate in sub 1mm tracking areas.

Previous studies have shown that with Simi’s markerless motion capture procedures, almost identical data to marker-based capture was collected for segCoGs and CoG over all recorded motions; and data correlated highly for segment velocities, segment accelerations, and joint angles in different joint movements compared with marker-based capture. The studies show that markerless MoCap has higher test-retest reliability and objectivity than marker-based MoCap as the 3D model will not change for a given athlete whereas marker placement may change or skin movement artefacts may occur due to movement of soft tissue or clothing in marker-based capture.

Marker-based tracking lacks in accuracy when compared with “gold standard” methods such as fluoroscopy, Roentgen photogrammetric techniques and intra-cortical bone pins (Stagni, et al., 2005, Maslen & Ackland, 1994; Tranberg, & Karlsson, 1998; Fuller, et al., 1997; Benoit, et al. 2007; Reinschmidt, et al., 1997). Simi’s hardware and 3D models/algorithms have improved drastically since the previous studies linked, they proclaim that they now attain “gold standard” accuracies, unlike marker-based systems.