PDA

View Full Version : Summary: Is it possible to marry kinetic and kinematic data inVicon or BodyBuilder?



Sarah R. Sullivan
06-16-2004, 05:08 AM
Sorry this one took me a little longer to reply to. I am currently
still working on this project. If anyone has any updates, please
don't hesitate to contact me.

Thanks to all those who replied.

Sarah R. Sullivan
PhD Candidate
Rutgers, The State University of New Jersey
Piscataway, NJ 08901
sarsulli@eden.rutgers.edu
908-420-3371


1.)

Hi Sarah,

I'm assuming that you're using C3D files - we offer an application
called C3Deditor that might be able to marry the two data sources
together. You should be able to open the Vicon generated C3D file
and "import" the kinetic data via ASCII - although I doubt that the
procedure will be quite this simple.
Alternatively, another product called C3Dserver, allows you to
write simple programs (in Visual Basic, Excel macros etc) that could
do what to want - this would be a better approach if you plan to
import data on a regular basis.
Both programs are available from our web site (www.emgsrus.com) or
the C3D web site (www.c3d.org) - I'd be happy to talk to you about this
when I get back in the office (next week) if you're interested.

Regards,
Edmund Cramp
Motion Lab Systems, Inc.
15045 Old Hammond Hwy, Baton Rouge LA 70816
Tel: +1 225 272-7364, Fax: +1 225 272-7336


2.)

Sarah,

You can do this with the AnyBody Modeling System. We recently assisted
a graduate student at Rutgers in doing this. If you know him, you may
want to contact him and ask him about it. Otherwise, please check out
the following sites:

- www.anybodytech.com
- www.anybody.aau.dk

Best regards,
John Rasmussen, Assoc. Prof., Ph.D.,
Institute of Mechanical Engineering,
Aalborg University.
www.ime.aau.dk/~jr jr@ime.aau.dk Tel. +45 9635 9307


3.)

Dear Sarah:
I am a Taiwan PhD student. Several month ago, I tried
to used body builder to calculate the kinetic data of
pitching. I also got some troubles and these troubles
still unsloved. There are some questions and reply I
asked Vicon and Vicon replied me as below. I hope
these information may help you.
~Good luck ~
Yu-Chuan Liu.

------------------------------------------------------
1.After reading the golem.mkr file, I didn't
understand how did it to calculate kinetic parameters,
used inverse dynamic or the other way.

2.Lately, I did some research about throwing. In my
study, Kinetic parameters were needed. But in the
golem.mkr file, the ball or the other objects to be
projected was not included. If I used the golem.mkr
file to calculate, the kinetic data still been output
although the golem.mkr didn't include ball. Was these
kinetic data correct? Because in inverse dynamic, ball
should be contained in the formula to calaulte kinetic
data. If using golem.mkr to compute wasn't proper,
could you offer the suitable mkr file or where could I
find the proper mkr file?

Dear Yu-Chuan Liu,
I am attaching a .pdf copy of the BodyBuilder manual.
This includes details on both of your requests.

Useful information on Kinetics in BodyLanguage
modelling can be found on pages 76-77 and 103-108.

You can also read sections on the connection of other
objects to the hand on page 77, though you may want
simply to add the mass of the ball onto the mass of
the hand using the anthropemetric definition of the
hand mass in the 'AnthropometricData' section of the
Golem model.

It should be noted that the Golem model supplied as an
example model does not include Kinetic calculations
for the upper body, only for the lower body and so
would need to be modified to include upper body
Kinetics.

Regards
------------------------------------------------------
Another questions are relating to euler angle and
kinetic calculation in bodybuilder. More details about these questions
are as below.
1. Few months ago, I mailed to you to ask the theorem bodybuilder used
to calculate kinetic parameters like torque, force etc. You mailed some
manuals about bodybuilder and I read them. Consequently, I couldn't
find any explaination to interpret the theorem in which vicon used to
calculate kinetic parameter. The manuals merely told me you can read
related textbook to find more details about kinetic formula. Would you
show me more details or reference and let me understand clearly about
the way bodybuilder used to calculate kinetic data?

2.The smae question like manuals to explain the way body builder to
calculte euler angle. In engineering textbook, I found the author say
if you want A coordinate system to rotate and reach the place B
coordinate system locates, both A and B coordinate system must have
one axis to parallel each other. But in the manual of bodybuilder,
the rule was not brought up. So I want to ask that--Does the body
builder follow the rule or not---.

3.The manual about bodybuilder also say something about floating axis
and fixed axis. Are there any other references about them.

Hi Mr.Liu,
I hope those are helpful.
Regards,
Stanley

Title: How does the FORCEVECTOR BodyBuilder macro work?
URL: http://support.vicon.com/cgi-in/vicon.cfg/php/enduser/std_adp.php?
p_faqid=383&p_created=1026913020

Title: Kinetic extensions to BodyBuilder
URL: http://support.vicon.com/cgi-in/vicon.cfg/php/enduser/std_adp.php?
p_faqid=451&p_created=1043965477

Title: Body Builder Sample Models
URL: http://support.vicon.com/cgi-in/vicon.cfg/php/enduser/std_adp.php?
p_faqid=507&p_created=1053111316


Dear all:
I am a graduate student in Taiwan. Recently, my lab brought Vicon
system and it's calculating software--bodybuilder. After some trying
and practice, I got some questions. The details about my questions
were as below.

1. Is there anyone ever using bodybuilder to analyze throwing or
pitching? In the bodybuilder, euler angle was used to calculate joint
angle, but wrist angle (containing palmar/dorsi flexion, ulnar/radial
flexion) elbow angle (containing extension/flexion,
pronation/supination), shoulder angle (containing horizontal abd/add,
abd/add, internal/external rotation) and trunk angle (containing trunk
rotation) were so complicated that I got some trouble to write
the mod file mp file and mkr file.

2. I couldn't find any explanation to interpret the theorem in which
vicon used to calculate kinetic parameter. The manuals merely told me
you can read related textbook to find more details about kinetic
formula. Does anyone can tell me more details or reference and let me
understand clearly about the way bodybuilder used to calculate kinetic
data?

3. To calculate euler angle, floating and fixed angle were used.
Although the manual of builder list some interpretations, I can't
comprehend them very clearly. Does anyone can tell me more details or
reference about this questoin?

Response (Dudley Tabakin) 12/04/2003 04:06 PM

Dear Tim,
1. Yes there are a number of people using BodyBuilder to analyse
pitching and throwing. In fact if you have a look at our download site
you will find a specific model written by a customer that looks at
baseball pitching. The model is called Baseball and can be downloaded
from our website. I also suggest that you download a model called
TrainingKin which is an upperbody model with complete Kinetic and
Kinematic calculations. The URL is
http://www.vicon.com/main/support/downloads.shtml

2. Here are some references for how inverse dynamics is done. The
Bodybuilder manual does give all the information necessary about how
BodyBuilder calculates Kinetics. The inverse dynamics procedure on the
other hand is not discussed as this is a well known, validated and
used technique in Biomechanics:

References:

Davis, R., Ounpuu, S., Tyburski, D. & Gage, J. (1991). A gait analysis
data collection and reduction technique. Human Movement Sciences, 10,
575-587.

Ramakrishnan, H.K., Masiello G. & kadaba M.P. (1991). On the
estimation of the three dimensional joint moments in gait. 1991
Biomechanics Symposium, ASME 1991, 120, 333-339.

Winter, D.A. (1990) Biomechanics and motor control of human movement.
John Wiley & Sons, Inc.

3. The manual has a very good description of Floating vs fixed angle
calculations in BodyBuilder. If you want to know how the angles are
calculated then it is best to referto a Mathematics textbook with
explanations of Euler angles.

I have attached a training document to this email that does a
simplified explanation of most of the concepts involved in modelling
and using BodyBuilder. Please have a look through this and the models
that you have downloaded from our website. This will give you all
the information you need.

Regards

ycl110@yahoo.com.tw


4.)

Sarah,

The most obvious solution is to feed the output of th SmartWheel into
the A/D board of the VICON (alongside the forecplates), ask the VICON
guys for support in doing this. This is so obvious though that I'd
guess you don't have an A/D board in your VICON system, ask your
supervisor to buy it and tell him the project isn't viable without it.
Small errors in synchronisation between kinematic and kinetic data can
make significant differences in inverse dynamics (there are papers out
there looking at slight offsets in sampling between video and force
measurements). I thus think you have to make sure you have an reliable
way of synchronising the signals (i.e. I think you will have problems
if you try a quick fix like solution).

Richard
richard.baker@rch.org.au


5.)

Hi Sarah

For the necessities of our study we managed to marry the data from a
Cybex Isokinetic dynamometer with the Vicon motion system. We just
connect the analog outputs from the Cybex with the vicon box that you
collect all the analog data.

Then we set up the analog channel into the vicon workstation software,
we set the proper scale factor for each analog channel and it was
working perfect. I compared the kinetic data that we collect with
Vicon with these from the cybex system and it was exactly the same.

For your question about the calculation of inverse dynamics I can't
answer to you because I didn't try yet to do this calculation with
Vicon softwares. If it is possible it will be very helpful also for
me.

If you have any further questions about the set up that we used,
please do not hesitate to contact me.

Dimitris Tsaopoulos
PhD Student
Institute for Biophysical and Clinical Research into Human Movement
(IRM)
Department of Exercise & Sport Science
Manchester Metropolitan University
Hassall Rd Alsager ST7 2HL
Great Britain
Tel: (+44) (0) 161 2475427
Email: D.tsaopoulos@mmu.ac.uk


6.)

You need to be able to write your force data into the C3D file as a
valid force plate type.

b.a.macwilliams@m.cc.utah.edu


7.)

Hi Sarah,

The best place to go for this type of informaiton would be the Vicon
support department. You can call them at 949-472-9140 or go to thier
website at http://support.vicon.com.

I used to work in the support department at Vicon. I will do my best
to propose a solution but I would double check it with Vicon since the
software may have changed since I was there last.

In order to read non-forceplate analog data into BodyBuilder, you need
to collect the data as if the analog device were a forceplate. This
means that in the analog setup window, a force plate is declared and
the channels associated with each force vector are really assigned to
the channels where the data is collected. Naturally, you can still
set scale factors as appropriate for the device. This process just
stores the data in the force plate section of the c3d file.

If you have already collected the data, as I suspect you have, then
you can open each individual trial and changes the analog settings and
force plate set ups for each individual trial. I have not had the
chance to test this, but as I recall, this is accurate. This step
does not change the raw data that is collected, it merely changes how
the data is stored in the c3d file.

Once thiis is done you then need to make some accomodation in BB. To
call the data, you will need to compose a rection. Remember that a
reaction consists of force, moment, and point of application. For
example:

ExternalDevice=|forceplate1(1),Forceplate1(2),Forc eplat1(3)|

You can then create a reaction with a segment, i.e.

RHand- ExternalDeviceSegment,Rhand

RWrist=Rhand,RForearm

etc, etc

Then you can create a reaction by saying:

ForceonHand=Reaction(RHand)

and then of course, you can decompose the components of the reaction
and then output them.

I will give some disclaimers here:

1. BB may have changed since I last used it, check with support for
confirmation

2. I do not have any manuals/software available, all suggestions are
from memory...when in doubt refer to manuals.

3. Check for the most recent versions of the software, including Plug
in Modeler. There may be a plug in that performs this action and
makes your life a lot easier.

4. I have made the assumption that the device sampling rate is a
multiple of the camera sampling rate. If it is not, then you have to
incorporate some re-sampling algorithm.

Another option would be to contact http://www.vaquita.co.uk. They
will create custom plug ins for you, for a fee. The developer at this
company used to be part of the development team at Vicon.


I hope this helps. Let me know if you have any questions.

-Michael Conuel

mconuel@hotmail.com


8.)

Hi Sarah,

If there's an analogue output from the SmartWheel, you should be able
to hook it up to the Vicon analogue input, to one of the forceplate
channels. The data will then appear as a Forceplate in BodyBuilder.
Your only worry is in getting calibration correct.

Failing that, if you have access to the SmartWheel data files and can
get it into the right ASCII format, you could import it into the
motion capture trial using the ASCII Import PlugIn which is on my
website(www.vaquita.co.uk). You could import the data either as a
reaction in the form of trajectory data, or again as forceplate data.
It makes little difference as far as BodyBuilder is concerned. Here
the only problem would be synchronisation.

Certainly once you get the data in there, BB will be able to hanlde the
inverse dynamics. Though you may have to be clever about how you
handle the reference point, and you will need to use the "Connect"
function to connect the reaction to you hand segments. I guess the
SmartRim delivers a simple torque - you will need to manipulate this
in the BB code to give you a reaction with a reference point - e.g.
the rim axis (centre).

Well, I hope these pointers get you going - I can help further if
needed...

Timothy Pitt

Vaquita Software www.vaquita.co.uk
Lets you concentrate.

tim@vaquita.co.uk

-----------------------------------------------------------------
To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l
Please consider posting your message to the Biomch-L Web-based
Discussion Forum: http://movement-analysis.com/biomch_l
-----------------------------------------------------------------