simtk.org

Thanks deeply!I'm using the toolkit and I can't help to say this is really great!

I encountered two problems now, any help would be thankful.

1. When running the testWalking, the toolkit could not continue when showing "the file testWalking_myEMG.mot could not be read", I don't know why because that file has been generated at that time.

2. I really got confused about the coordinate system transforming formulas. As you mentioned in the user's manual, "

% This section sets transformation vectors that are crucial to the
% rigid body transformations needed between the Force Plate, Vicon,
% and the Model coordinate systems.
%
% Example: FP coord system -> Model coord system
% glab.dirVec.FPMODEL(3,:) = [2 -3 -1];
%
% MODEL X = FP Y
% MODEL Y = FP -Z
% MODEL Z = FP -X

then in my thought, 2 lines of that are enough for the transforms among 3 coordinate systems, why need 8 lines like below in the "LOADLABLES.M"?

% X direction (Vicon) Gait -> transformation vectors
glab.dirVec.FPMODEL(1,:) = [-1 -3 -2]; % FP coord system -> Model coord system
glab.dirVec.VICMODEL(1,:) = [1 3 -2]; % Vicon coord system -> Model coord system

% -X direction (Vicon) Gait -> transformation vectors
glab.dirVec.FPMODEL(2,:) = [1 -3 2]; % FP coord system -> Model coord system
glab.dirVec.VICMODEL(2,:) = [-1 3 2]; % Vicon coord system -> Model coord system

% Y direction (Vicon) Gait -> transformation vectors
glab.dirVec.FPMODEL(3,:) = [2 -3 -1]; % FP coord system -> Model coord system
glab.dirVec.VICMODEL(3,:) = [2 3 1]; % Vicon coord system -> Model coord system

% -Y direction (Vicon) Gait -> transformation vectors
glab.dirVec.FPMODEL(4,:) = [-2 -3 1]; % FP coord system -> Model coord system
glab.dirVec.VICMODEL(4,:) = [-2 3 -1]; % Vicon coord system -> Model coord system

Anyone mind sparing a little to explain? Thanks!
Yi.

The answer to 1) could be something like incorrect file name or file does not exist. check that all the steps have been executed properly and the fileames are correct.

As for 2), this all depends on your lab configuration. I have assumed (as with most vicon setups) that the Z direction is always vertical. Hence the horizontal directions (the possible vicon directions for gait are +x, -x +y, -y). Since we are mapping the model coordinate system to the vicon coordinate system, you must know in the trial which way is forward and backward (the forward direction of walking will therefore correspond to the +X direction in the opensim model). so for example, if the force plates are layed out in the +/- Y vicon direction and you are walking forwards in the -Y direction, you want to select direction=-2. If you did select direction=+2, then the toolbox would think that the person is walking backwards. I hope this makes some sense.

Cheers,

Tim

Thank you so much.

For the Q1, I have found the problem and file was missing so it solved.

For the Q2, I though I thought I have understood the coordinate system transform in the toolkit, but I still could not extract my data at the right direction.

So let me first express the configuration first and at last I will upload my data at a ZIP folder so if it would not spare you too much time would you mind checking it where was wrong?

First this picture I took my cell phone which is in my thought I recovery your coordinate system on the left and create my coordinate system on the right. Sorry for the blur picture.

http://www.wuliaole.com/downloads/gaitt ... dinate.jpg

So if you can read this, I set:

glab.FP.prefix = {'CH02','CH03','CH04','CH05','CH06','CH07'};
glab.FP.suffix = {'','','','','',''};
glab.FP.verticalForceIndex = 3;

Here is my first question, the number 3 here represent the 3rd channel? I mean here in my analog the 'CH04' is force on Z-axis, so am I right? And another question about this is that the vertical force here should be the coordinate system of VICON system, I mean the coordinate system in the bottom of three coordinate system?

Then I was wrong about this, please figure out. I have spending several days on making extracting data correctly but I think it's still far away.

Then here is my second problem. If my coordinate system was right, so my subject should take the +x as the forward direction, so I set:

keyDyn = getEvents(c3dFileDynamic, 1, 1);

Because I only got one force plate, so I think the first '1' represent the +x forward, and the second '1' represents the only force plate? right?

Then I thought that's all, I uploaded a zip file including the pictures of the results, but the subject is taking +x direction but ground reaction force seems taking the '+z' as the forward direction in OpenSim or the motion generated by your toolkit.

http://www.wuliaole.com/downloads/gaitt ... _40c3d.zip

Tim, thanks for your reply, I kind of felt weird because I am the only person who post topic here, so I can help of thinking others got no questions about your amazing toolkit, which is embarrassed me to some extent. But I really hope to extract my data properly because I got hundreds of data on my hand, so hope not to bother you too much.

At last is my raw data C3D file and TRC, COORDINATE, KINETICS files extracting from the C3D file. Please help.

http://www.wuliaole.com/downloads/gaitt ... T_LF_5.zip

Thanks in advance.
Yi.

Hi Yi,

Having a quick look at the c3d file, the markers/analog channels dont seem to have any naming convention. i.e. the marker names are '0000', '0001', etc. This does make life a lot more difficult and I think you would need to rename the channel names so they follow a better convention. Particularly for the force plate analog channels, you will need to have names with meaning (i.e. Fx1, Fy1, Fz1, Mx1, My1, Mz1 rather than CH02, CH03, CH04, CH05, CH06, CH07). You can rename the channels in Vicon. Let me know if you still have problems.

Cheers,
Tim

Also take a look at http://www.youtube.com/watch?v=jwVc6e_wlgM

Hey!

I've tried to use the toolbox, but it keeps displayins error when using the function "getEvents"....the message says the forceplate's type is not supported!!

Any idea how can I solve this?

Thanks,
Marta

Hi Marta,

Perhaps you could send me your c3d file and loadlabels.m file that you are using (t.dorn@pgrad.unimelb.edu.au), and I'll take a quick look to see if I can identify your problem.

Tim

Hi Yi,

I had a look at your kinetics file and was interested to know what events you used to split the grf's between the two feet? I was confused as only one foot strikes the plate yet you have split the data between the columns for the left and right foot? Am I interpreting this incorrectly?

Sam