PDA

View Full Version : Question replies...



Jason Lake
07-25-2003, 07:47 AM
A big thankyou to anyone who took the time and effort to reply to my
question. Below is a list of the responses that I recieved. Again,
thankyou very much.
Yours sincerely, Jason Lake, University College Chichester, UK.

Dear Jason,

the most simple way to do that is to record the same signal (an
electrical
signal modification for example) both on forceplate acquisition device
and
on cameras (a light which will switch on or off) ....

I'm also interested in the summary of replies ...

Hope it helps ....

Guillaume Rao

Use a clock signal simultaneously triggering an LED (vide will pick
this up) and being sampled by the ADC spare channel along with the
force plate.
Kind regards
a

Dear Jason,

we have solved a similar problem using a LED coupled with vertical GRF.
The
LED went on when stepping on the force plate and was visible in all
cameras. The problem is only, to make it visible in all cameras at the
same time. Another problem might be the sampling rate, i.e. if the
force plate measures with 1000Hz and the cameras with maybe 50Hz. It┤s
helpful, if
your
cameras are also synchronized, i.e. they start at the same time.
Otherwise
you have a time delay of 0-10ms...

I hope, I could help you.
Good luck,
Hagen

Hi,

If you have en extra channel available in the device you use for
collection of force platform data, the simplest approach is to add a
switch which generates a pulse on the free channel and lights a LED
visible on the video image.

- Marko

Dear Jason,
A simple method to synchronize your video and force would be to get an
low voltage light emiting diode (LED), connect a thumb switch to
open/break the circuit on depression, place in series a 9 volt or (2)
1.5 volts battery holder, and then terminate the circuit with a BNC
connector or whatever connector your A/D board for the force plate has.
Whether you use a 9 or 3 volt battery depends on the voltage input
range of your force plate A/D. A switch that drops the voltage tends to
work faster than a thumb switch that closes the circuit upon closure
and the voltage has to build up and a square wave is produced on your
A/D channel and the LED is simultaneously seen in the camera. The LED
actuates quickly and normally can be seen from several cameras and this
device will cost at most $10 (US) and requires just a soldering iron,
and electrical tape.
Good luck!
Al Finch
Biomechanics Lab
Indiana State University

Dear Jason,

Here at Experimental Zoology,Wageningen University, the Netherlands we
use a random pulse generator to synchronise videodata with EMG or other
data. We send the pulse both to a LED (visible in the video) and to the
data-aquisition board that is aquiring the other stuff. After that it's
just a case of matching the two datasets. Good luck!

Jochem

The very simplest way is for someone to tap the plate with the nd of a
cone
or some such implement. This produces a spike obn the vertical force
trace,
and the moment of impact is visible on the video. The two sets of data
can
then be matched up using Exel or some such spreadsheet system. Hope
this helps Su Division of Psychology and Sport Sciences, Northumbria
University, Wynne-Jones Centre, Mewcastle Upon Tyne,, NE1 8ST Phone
0191 227 3712

Hi Jason,

First, are your video and force plate collections started by the same
trigger (dependent on your system)? If so, then if there is a synch
discrepancy it should be the same each trial (you'll want to check
this) and
can be easily corrected by software/post-processing.
An easy method to check the synch is to add reflective tape to a golf
ball
and drop onto the force plates. The lowest Z (vertical) coordinate of
the
golf ball should correspond to approximately (depending on your sampling
rate) the radius of your golf ball. Your force plate transient should
occur
between this video point and the next one in time. To be more
accurate, using your video data you could calculate the exact time the
golf ball
hit
the ground and match this up with the force plate data at that instant
in
time. Sampling at higher rates obviously increases your accuracy.

Hope this helps,
Matthew Walker
Motion Analysis Laboratory
Shriners Hospital for Children, Erie, PA
mrwalker@shrinenet.org

Hi Jason,

Take a look at the following references:

Bray, K. and Kerwin, D.G. (2003). Modelling the flight of a soccer ball
in a
direct free kick. Journal of Sports Sciences, 21, 75-85.

Kerwin, D.G. and Trewartha, G. (2000). Strategies for maintaining a
handstand in the anterior-posterior direction. Medicine and Science in
Sports and Exercise, 33, 1182-1188.

They used a pair of LED arrays to sync. their force plate data with
kinematic data. The kit is available from WeeBeasty Electronics in
Loughborough and I can forward on the contact details to you if you are
interested in getting hold of some.

Cheers,

Paul Glazier
Senior Sports Science Technician (Biomechanics)
UWIC

Dear Jason,

SIMI Reality Motion Systems is the leader in developing video based
motion
analysis software. We are able to capture signals from up to 12 DV
camcorders together with a forceplate and/or EMG directly to the
harddisk.
Both systems are synchronized with an accuracy of 1ms. Just press one
botton
and acquire data from a multiple camera setup and your forceplate
without
any external signals.
We successfully equiped some institutes in the UK like Manchester,
Leeds,
Sheffield, London...

Feel free to contact me in case of any additional technical or
financial questions Best regards Stefan

!!! You can SIMI in Motion !!!

Stefan Klippel
Sales Management Motion Systems

SIMI Reality Motion Systems GmbH
Max-Planck-Str. 9-13
D-85705 Unterschlei▀heim
Germany
+49 (0)831/6971752 (Office)
+49 (0)174/3154604 (Mobile)
+49 (0)89/321459-0 (Company)
+49 (0)831/6971753 (Fax)
mailto:klippel@simi.de
http://www.simi.com

Hi Jason,

Obe way to do this is to set your A/D to accept an external trigger.
Then you need some small circuitry to convert the output signal of the
video camera into a trigger pulse. You can use the vertical sync
signal of the video for this, then use a little electronics to convert
that video sync signal to a pulse for the A/D card. The problem is
that you are now sampling force at the video rate.


When we were building all of our own hardware we would generate
sampling pulses at say 960 Hz then send every 16th pulse to a vertical
sync generator, which triggered the cameras. That's a little more
electronics.

Good luck.

Gene Alexander
Stanford Biomechanical Engineering

Hi Jason
You can purchase synchronisation boxes from Kistler. From memory, you
put the box into the camera FOV and when you begin taking force
measurements, an light flashes on, allowing the commencement of the
force measurements to by sychronised with a particular frame. Have a
look around the kistler website- www.kistler.com
Cheers
Jackie

I forgot to mention, that was not my idea....someone (I forgot who)
from another lab gave me that idea

Michael El-Shammaa, M.S.
Engineer, Motion Analysis Center
Children's Memorial Hospital
2300 Children's Plaza, Box 92
Chicago, IL 60614-3394
773-327-1936

If you are collecting motion data with video and force, what I did to
check
synchronization was to wrap a golf ball with reflective tape (same type
as
on markers), and drop it on to the force plates while collecting video.
The
force should begin when the motion data is at its lowest point, and the
video should show this as the lowest point. It is expected to be 1
frame or
so off, depending on capture rates, as each of the 3 will capture at
different rates. I hope this helps? Best Regards,

Michael El-Shammaa, M.S.
Engineer, Motion Analysis Center
Children's Memorial Hospital
2300 Children's Plaza, Box 92
Chicago, IL 60614-3394
773-327-1936

Hi Jason,

You have to generate a signal which appears on the video and the
analogue
channels at the same time. You could build a simple circuit (battery,
LED,
switch, wire). This would light up the LED when you push the switch and
send
a signal to an analogue channel at the same time. Your force platform
output
is probably connected to an A/D board in a PC. That board is likely to
have
extra inputs not used for the force platform. Check the specification
of the
A/D board and generate the electrical signal so that you don't damage
the
board.

A related issue is that the video should run in synchron with the
analogue
acquisition. You probably have a solution to this already. If not then
this
is a more difficult problem and needs more investment. The details
depend on
your actual hardware.

Even if your analogue sampling and video are not exactly synchronised
you
can build the circuit but you have to declare that there is an inherent
delay due to lack of synchronisation between analogue and camera.

The elegant solution would be to trigger the analogue board to begin
sampling (through its digital inputs) and the camera to begin recording
by a
pushbutton. This would solve both issues above. Again, depending on the
camera and A/D board you use this may involve more advanced electronics.

I hope this helps,
Gabor
--
Dr Gabor Barton (MD)
Senior Lecturer in Biomechanics
The Research Institute for Sport & Exercise Sciences,
Liverpool John Moores University
Room 2.51 Henry Cotton Campus, 15-21 Webster Street, Liverpool, L3 2ET
Tel: +44 (0)151 231 4333/4321 Fax: +44 (0)151 231 4353
E-mail: G.J.Barton@livjm.ac.uk

One simple way is to drop a golf ball onto the plate just prior to the
action to be filmed. This way you can go back through the data and
find the
frame of video (50 Hz) and the force data (500 Hz?) where contact
occurs.
Easiest if you pick a multiple of your video frame rate as your force A
to D
rate. There are better ways like a voltage sync pulse, but none
cheaper.

Michael

Jason
typically in my lab we made a simple LED circuit that lights up when
the experimenter pushes a button. simultaneously, we monitor the
voltage across the LED with an extra A/D channel. so the square wave
that appears synchs with the LED coming on in the video.

don't use a light bulb, they actually take a while to become visible
after the voltage is on. the LED is nearly instantaneous.

rodger
--


Rodger Kram, Ph.D.
Associate Professor
Dept. of Integrative Physiology
Univ. of Colorado
354 UCB
Boulder, CO 80309-0354

Jason,

This may or may not assist you in your discovery. Attached is an
abstract that I presented. Hopefully, you will find an answer within
the contents. If not, email me personally and I will assist you
further. Thanks and good luck with your project. Please keep me
informed of your results.

Jeffrey B. Casebolt
CSU, Sacramento
Biomechanics Laboratory
6000 J Street
Sacramento, CA 95819

Jason,

I am assuming that your analog and camera systems are not
electronically synchronized (start and end data collection at the same
times). In this situation, we used a external light to synchronize
force plate and cameras. The light was triggered by the person
collecting data, and
sent a
electrical impluse to a channel (synchronization channel) in the analog
data collection system. The analog data collection system collected
force
plate data and the signal in the synchronization channel. During data
processsing, identify the first time in camera data in which the light
was
on and the first time in analog data the synchronization channel signal
was
high, and match these two times to synchronize force plate and camera
data.

We also used critical event method to synchronize force plate and
camera data in some situations. During data processing, we identify
the first time the foot is on the force plate and the first time the
vertical
ground
reaction force was greater than zero, and match these two critical
times to
synchronize force plate and camera data.

Force plate and camera data can also be synchronize with other methods.
Which method is the best for you depends on the specific systems you
have.
No matter which method you are going to use, the general principle is
the
same: you need a signal that is visible in both analog and camera data.

Bing Yu, PhD
Associate Professor, Director
Center for Human Movement Science
Division of Physical Therapy
CB# 7135 Medical School Wing E
The University of North Carolina at Chapel Hill
Chapel Hill, NC 27599-7135

Tel: (919)-843-8643
Fax: (919)-966-3678
E-mail: byu@med.unc.edu

In my thesis, I set up the camera and force platform speeds to be a
factor of the other. For example, my camera speed was at 1/120 hz while
my force plat was at 1/600 hz making it a 1 to 5 ratio. We used a synch
box connected to the force platform and also connected a LED light and
put that light in view of the camera. Using APAS while digitzing, I
noted the frame number in which the light was shown (usually best when
the light is activated slightly before the movement of interest) I then
synchronized my video frames to the force plate data (every 5th sample
of force platform data extrapolated) and thus synchronized my camera
data with the force plate data. The synch box was set up to
simultaneously start the force plate and the LED light together while
showing a voltage spike on the voltage data. This voltage spike
denoteds when the LED light was shown on camera. Once the LED light
video frame was matched with the force platform sample, (video frame
#12 = force platform sample #157 for example), then data was
synchronized with frame and force platform samples. Sorry if I confused
you even more! But just wanted to give you an idea on how I
synchronized my data.

Mitell Sison

Hi Jason,

I apologise for the delay in replying to your BIOMCH-L inquiry. I have
been out of town and recently returned.

Our company, Peak Performance Technologies, Inc. has a unit that was
made specifically for synchronization of video and analog sources,
which we call an Event and Video Control Unit. The especially good news
for you is that there is one on your campus in Chichester. Dr.
Rosemary Dyson has it. I imagine you could work something out with her
to borrow/lease it. If not, please contact me and I will see what
other options we could work out for you.

Regards,

Gary Scheirman
Vice President
Peak Performance Technologies, Inc.
www.peakperform.com

You can throw down the heavy mass on the platform and record it by
camera. This moment will be the origin of sinchronication.

Vera Talis, PhD
Laboratory of Neurobiology of Motor Control
Insititute for Information Transmission Problems,
Russian Academy of Science,
Bolshoy Karetny 19,
Moscow GSP-4,
127994 Russia
tel: 7 (095) 209-2895
fax: 7 (095) 209-0579

---------------------------------------------------------------
To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l
---------------------------------------------------------------