Motion Capture Flashcards
What is motion capture?
Process of recording anatomical landmarks and track its trajectories to perform kinematic analysis
What are the different types of systems in motion capture
2D, 2D/3D system Manual Digitizing, 2D/3D System Automatic Digitzing, Inertial Measurement Units
Other miscelleanous types of systems in motion capture
Radar gun, time gaits, acceleroemter, speedometer/linear encoder, electrogoniometer
What are the parameters measured in motion capture
Position, time, displacement, speed, acceleration
What are the pros of using motion capture?
Accurate and reliable
Less time consuming data acquisition
Enables immediate feedback through qualitative analysis
Minimal interference to performer
Can use in different environments (water, indoors, outdoors)
What are the cons of using motion capture?
Expensive Time consuming (analysis and handling) Requires dedicated technician with learning curve Software bugs and breakdowns Large dataset to handle
Procedures for Motion Capture? What are the differentiating factors.
2D or 3D kinematics
2D, one camera and coplanar. Calibration frame or pole with minimum 2 points.
3D, two or more cameras and non-coplanar. Calibration volume, L shaped + T shaped calibration kit. Minimum 6 non-coplanar points (recommended 15-20).
What are the suggested sampling rates for different activities? What is the theorem that it’s based on?
Nyquist-Shannon sampling theorem
25-50hz: walking, swimming, stair climbing
50-100: running, shot put, high jump
100-200: sprinting, javelin throw, football kick
250-500: tennis serve, golf swing, fencing parry
Usage of markers in motion capture, how are they used? What types?
Placed according to anthropometrical model CoM segmentation method.
No marker, passive marker (reflective), active marker (emitting light)
How are markers captured or tracked?
Manual (frame by frame)
Semi automatic (kinovea)
Automatic
What are the concerns when doing manual or semi-automatic marker tracking?
Use same operator to digitize all trials, ensure consistency
Consider anatomical landmark being sought
Calibration frame / control points should be carefully digitized
Check that 3D reconstruction errors fall within acceptable limits after completion of calibration
Representative sequence should be digitised several times to establish intra-operator reliability.
How are 3D motion captures calibrated?
Coordinate transformation using Direct Linear Transformation (Abdel-Aziz & Karara 1971), a 3D reconstruction algorithm.
Describe the considerations when transforming coordinates in a 3D construction.
Calibration frame > large enough to include space of motion (too small, danger of extrapolation and hence inaccurate coordinate computation)
Don’t alter camera settings after calibration is done
Include as many control points as possible
Use as many cameras as possible (reduce redundancy for reconstruction)
Set control object (calibration frame) properly to align axes well in relation to direction of motion.
How is data from Motion Capture processed/analyzed?
Apply low pass filter
Determine cut off value - through residual analysis
Optimal cutoff frequency > one where residual starts to change very little (this point considered as filtering mostly noise and minimally signal)
Export the raw data (coordinates+time)
Filter the signal after residual analysis (on acknowledge: transform digital filters IIR
Low pass)
Compute the resultant speed (vR2=vx2+vy2)
Plot the speed-time curves (data displayed vs. data filtered) on a graph:
What are the differences identified between 2D Video Analysis (manual and tracked) and 3D Optoelectronic systems?
Knee angle measured had few diff between systems
3D model produced larger angles at midstance
Possibly due to how 3D model locates hip joint and addition of marker clusters
2D videography similar results to 3D model when manual digitizing as it allowed for skin movement errors to be corrected
