Intrumentation for Motion Analysis
sampling rate used to be 4 photos per second.
The human eye sees 10-12 frames/sec. Any thing faster appears to be smooth movement.
Home video cameras are 30 frames per second. This system scans through the data two times.
Now we have the capacity to video at 60 frames per second and even relatively
inexpensive equipment can go up to 200 frames per second.
Might be up to 40,000 Hz ( frames per second) to measure impact studies such as
high speed car crashes or materials testing for cycling helmets etc.
The faster the motion, the faster the frame rate needs to be.
Want a high enough frame rate to be able to capture whatever the most important
high speed event is in the movement.
Sampling theorem states: you want the smapling rate at least two times as fast as the
fastest frequency. The fastest frequency is determined by a fast fourier analysis to
find a frequency spectrum of the movement
Example: impact for a golf swing or tennis racquet requires a faster sampling rate than heelstrike during running or walking.
When we digitize we convert analoque signals to digital data (X,Y,Z) coordinates,
from there we calculate displacement, velocity, acceleration and angles etc.)
With our system you can film and use:
1) light reflective markers on a dark surface
2) dark markers on a light surface
If there are markers missing, data can be entered by hand.
With our system we time encode the video on one of the two sound channels and
the computer can keep track of the beginning and end of data or important points in
Our system picks up 16 frames of data at a time - processes it and then picks up another 16 frames.
Some systems can digitize in REAL time but the data setup process has to be perfect.
Our system is capable of 2 dimensional analysis ( height and width)
Some systems are capable of 3 dimensional analysis ( height, width and depth) but
you need to use at least two cameras and then the 2d coordinate data is converted to
3d coordinate data.
set up a spatial model- set up a stick figure
set up a project file- frame rate and spatial model name
Digitize- convert analogue to digital data.
Analysis- calculate the paramaters- disp, vel, accel, angles, ang vel, ang acc.
Convert pixel data to real values. determine how many pixels in some real distance
(a meter stick) and then convert the rest of the data.
filtering- makes the data smooth. Takes out digitizing error due to the machine or
jitter from the camera. Butterworth filter is a 5 pt moving average emphasizing the
current point the most and weighting the surrounding points less out from the
Can then calculate the data and plot whatever is interesting to you.
You can also pick out key variables such as maximum knee flexion during swing
phase and statistically compare groups of people.
General issues associated with Electromyography-
Surface electromyography to analyze the patterns of muscle activation.
Action potentials occurs slightly before movement actually occurs
Is earlier in phasic ( transitory ) than in tonic (static ) movements
Phasic shows more fast twitch activity tonic more slow twitch activity
magnitude of EMG proportional to the velocity of the movement ( also possible movement error in signal)
force is not directly related to the amount of EMG signal
EMG should be collected with video data at the same time to get joint information ( angles,
have to be careful when interpreting EMG data.
Heart rate as a means of getting a rough estimate of metabolic rate.
Heart rate and VO2 increase roughly in parallel so you can use heart rate as some indication
of metabolic rate.
Ground reaction forces- This can be collected to obtain ant/post, mediolateral, vertical
information about foot strike. To negate the effects of the shoe, sometimes this
data is collected bare footl
Force insoles can also be used to obtain force data at the foot. These generally only obtain
vertical ground reaction forces but they can collect data about the interaction of the foot with
Things to look for when reading a research article.
Information about data collection- methods should be repeatable.
1) a protocol section about subject training- to avoid training effect
2) subject information- age, height, weight, number of subjects
3) descriptions of the types of equipment- including model numbers
4) sampling rates for various types of data
5) statistical methods - qualitative versus quantitative