Welcome to the Podiatry Arena forums

You are currently viewing our podiatry forum as a guest which gives you limited access to view all podiatry discussions and access our other features. By joining our free global community of Podiatrists and other interested foot health care professionals you will have access to post podiatry topics (answer and ask questions), communicate privately with other members, upload content, view attachments, receive a weekly email update of new discussions, access other special features. Registered users do not get displayed the advertisements in posted messages. Registration is fast, simple and absolutely free so please, join our global Podiatry community today!

  1. Have you considered the Clinical Biomechanics Boot Camp Online, for taking it to the next level? See here for more.
    Dismiss Notice
Dismiss Notice
Have you considered the Clinical Biomechanics Boot Camp Online, for taking it to the next level? See here for more.
Dismiss Notice
Have you liked us on Facebook to get our updates? Please do. Click here for our Facebook page.
Dismiss Notice
Do you get the weekly newsletter that Podiatry Arena sends out to update everybody? If not, click here to organise this.

Analysis software

Discussion in 'Biomechanics, Sports and Foot orthoses' started by Ian Linane, Jan 18, 2008.

  1. Ian Linane

    Ian Linane Well-Known Member


    Members do not see these Ads. Sign Up.
    Hi

    Wondering if anyone has info on good software for analysis of gait/posture on CCTV? I'm aware of Vicon and Coda but I'm looking to see if there is software which I could apply to CCTV images whereby I can place markers on a person.

    Planning to email Southampton Uni next week to check with them as well.

    Thanks
    Ian
     
  2. If you are set on using cctv footage you probably need to look at a video based motion analysis system. I don't think Vicon or Coda would be suitable. You could import CCTV footage easily using something like Quintic etc. The key is getting the CCTV footage into your computer. You basically need to connect your CCTV video capture output to the computer video capture input and save in the correct format for the software. You might be able to firewire straight from the cameras. If not you could use a video card in your computer which is capable of importing via standard video connections. Back in the day, I used to do this with an ATI wonder card. Seems they still make something similar: http://ati.amd.com/products/radeon9600/aiw2006/index.html

    As I'm sure you are aware, such systems are 2D. The kind of footage you are likely to get from CCTV is not going to be suitable for measuring meaningful kinematic data.

    Hope this helps.
     
  3. Ian Linane

    Ian Linane Well-Known Member

    Hi Simon

    Thank you for the response. The CCTV material is generally brought to me on CD which I then put onto my Mac and work with.

    You are right that it is 2D and consequently has some significant limitations. In part this whole thing is an exploration to see if it has any value if combined with visual analysis of CCTV material and may lead to a project. It is something that has been one of those "I must get around to it things" and now I finally am. (might do an MSc in procrastination!)

    All input welcome.

    Thanks
    Ian
     
  4. Craig Payne

    Craig Payne Moderator

    Articles:
    8
    Silicon Coach
    Dartfish
    Trimfit
    Contemplas

    Get in touch with Jonathon @ Mar-systems for advice (he does sell Contemplas, but should also be able to give good advice).
     
  5. Ian Linane

    Ian Linane Well-Known Member

    Hi Craig

    Thank you. Will look at them all and contact Jonathon.

    Thanks
    Ian
     
  6. I which case you should have no difficulty importing it into a video based motion analysis system. I've often thought about potential application of cctv motion analyses in forensic podiatry. The problem is how do you get over the parallax issues? I guess if you had access to the "crime scene" geometry you could work it out by trigonometry.

    I'm intrigued, why are people bringing cctv footage to you for analysis? Do the "stars" of said cctv know they are being viewed by you?
     
  7. Ian Linane

    Ian Linane Well-Known Member

    Hi Simon

    "I'm intrigued, why are people bringing cctv footage to you for analysis?"

    A question I frequently ask myself. It began 7 years ago with a police enquiry. Has gone on from there and often involves defense teams as well.

    "Do the "stars" of said cctv know they are being viewed by you?"

    No. The material comes either from prosecution or defense. It will comprise of CCTV footage and/or covert footage that has been captured.

    As evidence, if it is used, it stands mainly as circumstantial. A link in a chain if you wish. Equally it has been used to eliminate people from enquiries.


    Craig:

    Very helpful contact in Jonathan, who sends his regards

    Thanks
    Ian
     
  8. So what data do you supply, i.e., what variables are you measuring?

    And a cheque for the banner at the top of the biomechanics forum.:rolleyes:

    Sorry Craig, I couldn't resist.
     
  9. Ian Linane

    Ian Linane Well-Known Member

    Hi Simon

    The data that can be collected is:

    1. Dictated from the outset by the level/type/quality of material provided. This is turn is dictated by the quality/types of images collected. Where possible I first view the original images. By the time a working copy of an image reaches me it will have naturally degraded to a small extent. Once work has been done based on this I ask to see the original again and compare what I have done against it.

    2. A further way in which it is dictated is in the type of movement available for viewing, the number of angles available to view the person/people from, quality of lighting, clothing surface etc.

    3. Types of data would include available images of limb rotation, trunk rotation, foot positioning, leg function, head and neck angles

    4. Work is always carried out to eliminate a person where ever possible, be the work provided by the police or defense.

    As an exercise it is very labour intensive

    The original presentation of this type of material in court was by Hadyn Kelly, a Podiatrist, back in 2000. Without being aware of this I undertook my first case in 2001. I believe Hadyn now lectures on this subject occasionally at Southampton Uni.

    Hope this goes some way towards answering the questions. You will have more questions than I can answer! I wrestle with some of my own!:bang:

    Ian
     
  10. How do you solve the parallax issues?
     
  11. Ian Linane

    Ian Linane Well-Known Member

    Hi Simon

    As I said: more questions than I have answers!

    With observational analysis of movement on CCTV I, currently, don't think that you can "solve" the issue of parallax. It is a given of the material you are presented with. It is a concern that I finally took to a forensic image analysis company. I worked through the approach I take with them with a view to critiquing it. To that extent the following are some of the ways that I try to work within it:

    1. Selection of the material you are willing to assess. Out of a half dozen cases you may
    preliminarily look at you may only do one because it is suitable to use.

    2. Ask for covert footage to be as close to the original as possible both in terms of height, distance
    etc. This cannot always occur and then you are left with other questions!

    3. Ensure that in the review process you are seeing what you thought you saw; reconsider
    influence of angles

    4. In the report be as (what the court calls) "fair minded" as possible. This means acknowledging
    the limitations of the material viewed and as to why it is limited, addressing the arguable points
    in your report and giving rational and reasonable reasons for your opinion. It is surprising how
    often this does not occur in evidence presentation.

    5. Attempt a fail safe. No matter how strong comparability is, if there is one idiosyncratic difference
    that you cannot reasonable explain then suggest that this is enough to eliminate a person from
    the enquiry.

    I'm fairly sure that this does not answer your question satisfactorily but it is what I currently endeavour to work within. Other ideas and input welcome.

    One thing is sure Cross Examination works wonders on making your rethink!

    Ian
     
  12. Mart

    Mart Well-Known Member

    Hi Ian

    I was interested in this subject a little during my reading on gait analysis and dipped very briefly into the issue of gait biometrics. My impression was that the direction being taken by those in the field was more to do with sillouette pattern analysis than the looking at the typical kinematic parameters we might be intersted in as podiatrtists. That may of course allow your perspective to have some additional value in he overall picture. You may be already familiar with this approach but in case you are not I am attatching a little info, I have a copy of this paper in my library and will email it to you if intersted. good luck.

    Abstract—In this paper, we propose a new spatio-temporal gait representation,
    called Gait Energy Image (GEI), to characterize human walking properties for
    individual recognition by gait. To address the problem of the lack of training
    templates, we also propose a novel approach for human recognition by combining
    statistical gait features from real and synthetic templates. We directly compute
    real templates from training silhouette sequences, while we generate the synthetic
    templates from training sequences by simulating silhouette distortion. We use
    statistical approach for learning effective features from real and synthetic
    templates. We compare the proposed GEI-based gait recognition approach
    other gait recognition approaches on USF HumanID Database. Experimental
    results show that the proposed GEI is an effective and efficient gait representation
    for individual recognition, and the proposed approach achieves highly competitive
    performance with respect to the published gait recognition approaches.


    cheers

    Martin


    The St. James Foot Clinic
    1749 Portage Ave.
    Winnipeg
    Manitoba
    R3J 0E6
    phone [204] 837 FOOT (3668)
    fax [204] 774 9918
    www.winnipegfootclinic.com
     

    Attached Files:

    • GEI.jpg
      GEI.jpg
      File size:
      53.1 KB
      Views:
      141
  13. Ian Linane

    Ian Linane Well-Known Member

    Hi Martin

    Thank you. I would be very interested in reading the paper.

    I suspect that one major difficulty with applying some of the current biometric techniques (many of which are a foreign language to me!) being explored is that you have no control over the CCTV material you are given both in terms of distance of the figures, lighting, artifact element, clothing, terrain etc and distortions due to pixelation aspects, equally a person may naturally change aspects of their posture by where they place their hands. It is working with the CCTV material and is it at all possible to use any form of digitization on it together with VGA that presently interests me.

    As I said, it is something I have been meaning to get around to and am trying to do it now. At the moment all input is welcome. It may be that it is all beyond my very limited ability (and there are some brilliant minds working out there) but if folks are interested I will happily post any progress. Equally, within reason, I think its important to expose approaches to peers so thay can have fun picking it apart and increasing my hair loss;)

    Thanks
    Ian
     
  14. Mart

    Mart Well-Known Member

    I totally agree with this sentiment and likewise value the forum immensly to stimulate and critique ideas.

    I'll email you the paper, having had another look at it I realise why this topic for me was a very brief dip and not a languish.

    Attatched are a couple of links which I find somewhat "a spanner in the works" in terms of challenging how we normally try and explain those things we attempt to measure and the expainations we arrive at, important but somewhat perplexing considerations for those involved in gait analysis.

    http://www.univie.ac.at/cga/teach-in/psych/


    http://www.biomotionlab.ca/Demos/BMLwalker.

    the second of these sites is well worth a browse around other areas - it is a fascinating exploration.

    cheers


    Martin


    The St. James Foot Clinic
    1749 Portage Ave.
    Winnipeg
    Manitoba
    R3J 0E6
    phone [204] 837 FOOT (3668)
    fax [204] 774 9918
    www.winnipegfootclinic.com
     
  15. Ian Linane

    Ian Linane Well-Known Member

    Hi Martin

    Loved the "biomotion" one. Using the sliders is great fun.

    Cheers
    Ian
     
  16. Saw this and thought of this thread:
     

    Attached Files:

  17. Ian Linane

    Ian Linane Well-Known Member

    Hi Simon

    Thank you very much for this paper.
    You interest and input is appreciated.

    Ian
     
  18. Mart

    Mart Well-Known Member

    Ian just came across this in my travels which you may find interesting

    cheers

    Martin


    The St. James Foot Clinic
    1749 Portage Ave.
    Winnipeg
    Manitoba
    R3J 0E6
    phone [204] 837 FOOT (3668)
    fax [204] 774 9918
    www.winnipegfootclinic.com
     

    Attached Files:

Loading...

Share This Page