Motion Analysis Systems

[Dan Robbins]

Hi all,


In our lab we are currently considering expanding our 3D motion analysis system via increasing our cameras to a total of 8. Currently we have a Motion Analysis Corporation (MAC) system with 6 Hawk cameras operated via EVaRT software. MAC have advised that they do not sell the Hawk cameras, though a newer camera can be purchased and included in to our current set up.

However, I have noticed that for a comparable cost we could set up a whole new system if we switch the Optitrack system (http://www.naturalpoint.com/optitrack/), potentially creating a system with up to 16 cameras. Obviously completely changing the whole system is a fairly drastic move, plus the lower cost makes me a little dubious.

I am hoping that some of the biomch-L members will have experience of these systems and be able to offer some advice and/or feedback on whether there is much difference between MAC and optitrack systems?

[Matthew Chua]

Re: Motion Analysis Systems

Hi Dan,

From what I've been told, I don't think the Optitrack system has been used in a full blown clinical gait lab setting, although it has been used effectively in research settings. I have recently looked into the competing mocap companies as our lab is looking to do a full system upgrade. These days, almost everyone is making some version of a low cost camera with varying lower specifications, with Optitrack's cameras being the lowest cost. Vicon's Bonitas, MAC's Kestrel and Ospreys, and Optitrack's Primes are all largely comparable in my opinion.

However, as MAC has mentioned, they have an uncanny ability to mix and match cameras from all bands of their products, as it all connects to a central hub via the same ethernet type wiring. That allows for huge flexibility when you already have a well functioning high end setup, and just want to increase your capture volume or have some mobility in some cameras. With the other companies, you'll have to select all your new cameras from the lower range or higher range, with no mixing ability.

The other issue that I came across, and to me might be more important than the hardware capabilities, is the software that comes with the cameras. MAC and Vicon obviously have a more established and tested software experience with various packages and plugins setup, whereas Optitrack may not. So depending on how much processing or customization you do within the collection software, this may or may not be a stumbling block. So far, for a typical clinical gait setup, I really like the simple efficient setup of Nexus, but also like the customizability and transparency of Cortex. Unfortunately, I have no personal experience with Optitrack's software other than what I have heard from others.

[Dan Robbins]

Re: Motion Analysis Systems

Hi Mathew,

Thanks for the feedback, very useful information. Between checking out the hardware, acquisition software and various post processing software options, there doesn't seem to be enough hours in the day!

Regards
Dan

[Josh Baxter]

Re: Motion Analysis Systems

Dan,

Could you tell us what your application will be for these cameras? What type of spatial resolution are you hoping to achieve and what is your capture volume? Optitrack cameras seem to be slightly lower resolution in general compared to the competitors but simply increasing the number of cameras could drastically improve the performance, especially if marker obstruction is typical in the movements you hope to measure.

If you are studying human motion then the amount of skin artifact might be worth considering when buying these cameras. Some of these more expensive systems (high resolution cameras) might get you extremely good accuracy, but extreme camera accuracy is probably not that great once you consider the amount of skin artifact. I use cameras to track the motions of foot bones (cadaveric specimens) using bone pins and with that setup I am fairly comfortable saying that my accuracy after assigning coordinate systems and calculating joint angles is under 1 mm and 1 degree (probably closer to 0.5 mm and degrees).

One strength of Motion Analysis cameras is that they have zoom lenses, which make for slightly more forgiving setups. However, I do no believe that the lower-cost offerings from MAC provides this feature...

It might be worth asking for a demo. I'm guessing that all three vendors will be at the WCB this summer. They might even be willing to visit your lab space and give you a more personalized show.

I'm interesting to hear more about your application and the thoughts of others.

Josh

[Dan Robbins]

Re: Motion Analysis Systems

Hi Josh,

Thanks for your reply. I work in the department of sports therapy, therefore the primary use of the system will be analysing exercises technique and gait analysis. We have briefly discussed using the system to analyse cycling, though I doubt this will be our main focus. In addition, our colleagues from computing use the system for students projects on computer animation.

We are currently negociating with facilities about which room will be converted in to a new lab, therefore spatial and volume factors are hard to establish until I know the room dimensions.

Regards
Dan

[Kimberley Andersen]

Re: Motion Analysis Systems

Hi Dan,

I use an Optitrack motion capture system for biomechanical analysis at Adelaide University. We have an 18 camera system with Flex 3 cameras that we've had for the last 3 years or so. Hardware wise it's alright, it can be tedious to set it all up as everything it has no wireless but for the most part we haven't been able to tie any issues to hardware faults. Software wise, well this is another story. We haven't got the latest software (Motive), we're still running Arena, but it's not the greatest. Tracking of markers can be iffy so the post processing can take hours for only a second of data. It frequently crashes, particularly when setting up the cameras and making use of the greyscale image function to help align the cameras (actual video output so you can see what you're pointing the camera at). Cameras also tend to drop out but if you reopen the software they're there again. Motive is to my knowledge better, but I have no actual experience with it. Instead of Arena (once the cameras are all where we want them) we use AMASS by C-Motion and while it doesn't look as flashy it's much more stable. We've had minor issues with AMASS as well but most of the time these are minor and easily fixed. Others apparently running AMASS with Optitrack cameras haven't had any issues so I believe a lot of our problems are stemming from our computer rather than the software itself.
With regards to computer animation, Arena and I'm assuming Motive are geared towards that purpose. In Arena you can modify the data tracks (move, draw and delete data and swap tracks over). There are facial tracking options available and the skeletal body tracking would also probably be useful as well. AMASS is more research based in my opinion.

I've just reread what I've written and I feel like I'm bagging Optitrack a bit. The hardware is fine if a bit messy and the software issues may be fixed in the new program. If it hasn't been AMASS is a great alternative (although I like setting up the cameras in Arena then testing in AMASS), but you will need a different calibration wand and of course the software costs.

Hope this helps

[Meredith Evans]

Re: Motion Analysis Systems

Hi Dan,

I caught your post and some of the responses regarding OptiTrack cameras. I would not be too concerned about the price being low.

“… plus the lower cost makes me a little dubious.

I am hoping that some of the biomch-L members will have experience of these systems and be able to offer some advice and/or feedback on whether there is much difference between MAC and optitrack systems?”

We integrated OptiTrack on a client’s request about 2 years ago. Since then we have completed systems using Flex 13’s, Prime 41’s, and Prime 17W. The OptiTrack software provides little in the way of analysis but we are only interested in the quality of marker tracking as The MotionMonitor does all of the biomechanical setup and analysis. For marker tracking, we think the OptiTrack software performs very well if the number of cameras is appropriate for the activities being collected. The MotionMonitor synchs this marker data with other sources, some example collections are here: http://www.innsport.com/optitrack-videos.aspx. I would note that all of their new cameras are Ethernet based rather than USB. Ethernet offers a lot of benefits over USB and probably represents the future direction of their camera development.

This is not to take anything away from MAC. We integrate their cameras with The MotionMonitor as well and they have always produced excellent product.

Meredith Evans
mevans@TheMotionMonitor.com
The MotionMonitor by Innovative Sports Training, Inc.