PDA

View Full Version : Qualisys & Vicon motion capture systems: Responses



Bruce A. Kay
05-29-1998, 01:03 AM
Many thanks to all those who took the time to respond to my post. Only
those responses that directly addressed the post (viz., Qualisys & Vicon)
are in the summary.

Comment: This is a really great way to get in touch with all of the motion
analysis companies. I had calls from lots!
------------------------
Original posting:

We are in the market for a new motion capture system, for analyzing the
kinematics (force reconstruction not required) of gait and posture during
treadmill walking, and have found that Qualisys (MacReflex, ProReflex) and
Vicon appear to make products that suit our needs.

I have looked through the biomch-l archives and found some references to
these systems, but very little on their overall quality.

Does anyone out there have experience with either of these systems and
companies? In particular, we'd like to know how good the systems'
reliabilities have been and how good the companies' support of hardware and
software has been. Accuracy and ease of use (e.g., being able to quickly
collect a lot of trials and reconstruct the 3D data easily) are also high
on our list.
--------------------------
Received wisdom (lightly editted):
---
I have a colleague who uses a Vicon system for clinical evaluations:
http://www.santarosahealth.org/BioMechanics.html I saw his lab & system
recently and was very impressed.
----
I told him that I use PC Reflex (the pc version of mac reflex) and am
somewhat disappointed with the pc versions of their software. While the
staff at Qualisys has generally been helpful, we have had numerous problems
with bugs in the software, as well as some technical issues reguarding
accuracy when using redundant cameras. Although we have so far
successfully worked around all the problems we have had, we have spent
considerable time doing so. For example, I spent 2 days trying to figure
out calibration problems, only to find that it was an undocumented bug
where the system was expecting markers to be identified in a certain order
(it is fixed in later versions). So, while the accuracy of the system is
pretty good (for our applications), the software is easy to use as long as
things go well, but when there is a problem, it seems to be a BIG problem.
Again, I think this is mostly a problem of the software contractor.
---
I have an old VICON system based on PDP in my Lab(The Children's Memorial
Hlth Inst, Warsaw, Poland), but also I have worked with VICON Clinical
Manager (VCM) which is situated in SIU School of Medicine, Motion Analysis
Lab, Springfield, IL. I also worked with MacReflex in Univ. Coll. Hlth Sci
in Jonkoping, Sweden. From my personal experience I would recomend you the
VICON system with VCM. Although the price is higher than in the case of
MacReflex it is easy to use, has a very good software. You are able to
quickly collect the data, analyse them and have a ready report in about one
hour. It is especially important in clinical applications where sometimes it
is necessary to see the results while the patient is still present in the
lab. The newest software developed by MacReflex is similar to VCM but the
work with it was more difficult at least for me. Even labelling the markers
is easier in VICON.
---
We have a Vicon 370 system that we primarily use for bilateral lower
extremity gait analysis. I am extremely pleased with the system's
performance.
---
I have used PCReflex in our lab for about a year. And as with any new
equipement, there were mistakes and problems that were operator error. But
I feel that some of the errors could have been eliminated if there had been
better documentation of the software. The hardware is quite good but I
would rate the software as poor. The support staff at Qualysis are very
informed and helpful but they are more experienced with MacReflex.
The cost of calibration frames and markers are very high. In the work that
we do in our lab, it sometimes becomes necessary for each specimen to have
seperate markers. Therefore, having markers for each specimen can be
expensive. Also we want to have the maximum resolution possible for each
experiment. Their calibration frames are not cheap and do not allow maximum
resolution because of the lack of different sizes of calibration frames.
Therefore we make our own calibration frames.
I must say that they are trying to address all the problems I have listed
above. They are planning to release new software in the near future that
will eliminate alot of their present software problems. They also have a
new calibration technique using a "wand". But as far as I know there are no
cheap markers.
---
I can only respond to the PCReflex software that Qualysis developed. In short,
don't bother. This software package is drastically underdeveloped and in
constant state of beta. It comes with an excel add in that masquerades as a
kinematic analysis toolkit but is buggy and not so easy to derive useful
information. In essence you have an expensive data capture system with no
means of deriving kinematic information. Although the technical support
personnel have been extremely accomodating one should not have to spend more
time on the phone debugging software than collecting data. Oh by the way did I
forget to mention that there are NO manuals that come with the software
(hardprint or otherwise). You might have better luck with Mac version. Good
Luck.
---
It's been years since I've used the Vicon; I was mildly pleased with
it. I believe Gillette Children's Hospital Human Motion Laboratory (ST.
PAUL, MN) uses a multiple camera Vicon system (at least 4 cameras).
---
I have just read your message posted to biomch-l. I donīt know about
Qualisys. In the other hand I have been an user of Vicon for about 7
years - from an old PDP-11 based system to the VX (VAX based) system and
recently the PC based Vicon 370. This is a robust and user friendly
system. Capture and reconstruction are fast and automated. Good support
also. Maybe you will find costs high. There is a cheaper version of 370, the
140 system, that will probably suit your needs (this version doesnīt
work with force plates, for example). You may develop the software that
will process the data generated by the capture system or purchase their
clinical application - VCM (just for gait analysis - lower limbs). The
first option will be more or less
reliable depending on your biomechanical models, available time and
personnel and may give you more flexibility. There is also another
product (BodyBuilder) that may be customized to particular biomechanical
models.
---
I've been using MacReflex since last year. I'm willing to share the
information.
The reliability is quit good. Average reconstruction error for an
approximately 2m-cube was 1.3mm and the maximum error was 3.5mm. For the
cross-validation test using an approximately 2m*2m*1m caliblation frame,
not extrapolation, average error was 1.9mm. How do you think these results.
When the reliability test was done, I used three cameras which located 2.5m
above. Angles between camera1 and camera2 and between camera2 and camera3
were approximately 60 degrees. The following presentation shows the
validation of MacReflex including extrapolation: M.Levy and G.A.Smith:
Validation of MacReflex motion analysis system for three-dimensional
coordinate prediction. Proceedings of the XVth congress of ISB, Jyudskyld,
Finland, 1995. Their aftercare service is also good. When I asked some
questions by e-mail, the responses were very quick. Although I live in
Japan and the companies located in Europe or USA, the quick responses using
e-mail make me comfortable.
When I set up the system first time, communication error between Macintosh
computer and video processors occurred, although it was not big problem.
Once you complete the set-up, you can collect data very easily. A number of
trials to collect data depends on volume of hard disk.
---
We have used the Qualisys MacReflex System for our experiments and made good
experience with it. MacReflex was found to be easy to use (calibration,
framegrabbing, pre-analysis) and reliable. We used MacReflex for an
analysis of goal-directed movements of the upper extremities. Nearly all
the markers (we used 14 markers) were recorded continously. The only
problem occured if markers were occluded for a longer time. Sometimes the
system lost the marker completely and left it out afterwards, even if the
marker was visible again. For further analysis we use our own software and
statistic analysis software (Systat,SPSS).
---
I have a good deal of experience with motion systems. I have used
the Motion Analysis Brand system and the Elite System, in my previous
gait analysis work. Currently I am using the Qualisys system.
Prior to purchasing the Qualisys system I performed an
extensive search and evaluated all of the motion analysis
systems on cost, ease of use, service and accuracy. During that time
I found the Vicon personel difficult to reach and work with to
obtain information regarding their system - this lead me to believe
that I would also have difficulty obtaining service or technical
support. Another reason we did not choose Vicon is, at the time, the
Vicon system was significantly more money than the Qualisys.
Currently we are using a five camera Qualisys system
(Macintosh software) for monitoring people during seated movement.
I have not used this sytem for gait analysis. I've been using it for
the last 21/2 years and find it to be user friendly and reliable. I
I have worked with the techinical people at Qualisys and they have
always been responsive to my needs. If you run into problems,
typically new parts are shipped that same day. The longest we have
had to wait for a replacement part was a week. We have special
cameras with high intensity infrared lights and wide angle lens, and
they did not have this in stock, so they had to send for one from
their parent company in Sweden - but we still had a new camera in a
week. As far as accuracy of your readings, this depends on several things,
such as space size, target size and the accuracy of the calibration.
We have run a few tests and wit a 15'*15' space and 18mm diameter
targets and we are accurate within +2/-2mm. I work with the group in CT --
Andrew Mahar is our primary contact.
---
I have a 5 camera Qualisys system & have had it for about 4 years. I mainly
do gait, but also do hand motion, eye blinks, ergonomic task, & golf swings
I've used Peak, Motion Analysis, & Ariel systems in the past & I can say that
the speed at which you get 3D coordinates blows the rest of them out of the
water. (I mainly used older systems, but have travelled to other labs to
collect data on newer competitor systems & my statement still holds true).
A friend of mine here at the University has used Vicon & says the same thing
(Speed of collecting 3D coordinate data is faster). My system is Mac based &
I PERSONALLY don't like writing code on a Mac. For that reason I would look
at their newly marketed PC systems. Accuracy & reliability are not an issue
as I think they have the best hardware in the industry. If you can't do your
own programing for post processing then you may want to spend the extra $100K
for one of the competitors which has some post processing software included
---------------
Thanks to:
Vernon McDonald
Bryan Kirking
Maria Lebiedowska
greg_lange@srhcc.org
Jerry Alexander
Nelson Sierra
Ted Morris
Wolfgang Pagani
Jose Haroldo de Assis Cavalcante
tmatsuo@user.center.osaka-u.ac.jp (Tomoyuki Matsuo)
"T. Alexander"
"Tamara Reid Bush"
GSRASH01@ULKYVM.LOUISVILLE.EDU

Bruce A. Kay, Ph.D.
Assistant Professor of Research
Dept. of Cognitive and Linguistic Sciences
Box 1978, Brown University, Providence, RI 02912 USA
401-863-2608 fax 401-863-2255
Bruce_Kay@brown.edu personal webpage: http://www.cog.brown.edu/~kay/

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