lifting movement/ C3D

Hi,
I would like to use the GaitFullBody-Model to simulate a movement of upper body. therefor I use C3D-data which only contains the markers from head till pelvis. So what I have to change to run the model?
I have set the legs to 0 in BodyPartsSetup1/2 and have removed the unused markers from ModelSetup. I have also removed the Plate1-3 from EnvironmentAutoDetection.any.
Is this the right way? Or is wrong to remove these parts?
I’m a bit surprised that InverseDynamicAnalysisSequence still works, even if there is no marker created in ModelSetup.any.

Can you show me a general approach to this problem?

Thank you and best regards
Wolfgang

Hi Wolfgang,

Removing these parts sounds absolutely correct.

The analysis is running in two steps first you find the motion then this motion is used by the inverse dynamic solver.

If the inverse dynamic analysis is running and you have no makers defined in the modelsetup? then it might be running the motion from a previous analysis.

If you still have the markers from the upperbody present in the model then your approach should be ok.

Best regards
Søren

Hi,
I would like to use the GaitFullBody-Model to simulate a movement of upper body by using C3D-data which only contains the marker from head till pelvis.
Because of a limited number of cameras of our motion capture system, it was only possible to track 18 marker ( ‘LUPA’,‘LWRA’,‘LWRB’,‘LFIN’, ‘LASI’,‘LFRA’, ‘LELB’,‘LSHO’, ‘CLAV’,‘STRN’, ‘RSHO’ ‘RUPA’,‘RELB’,‘RFRA’ , ‘RWRB’ , ‘RWRA’ , ‘RFIN’, ‘RASI’). Now there is a problem with the mannequin folder, because “RPSI” and “LPSI” are not defined.In addition two markers jump back to the origin for a certain time during the movement.

1.Is there a posibility to load the model and run the studies with this limited markerset?

  1. Do you know a possiblity to ‘fill the gaps’ for the two jumping marker e.g. editing the c3d-file with a special software?
    I’ve tried to solve the problem like the tutorial did it in “Making things move, Lesson 6: Weight functions” but Anybody stops loading and outputs :
    ERROR(OBJ1) : C:/Dokumente…ellungen/l…0/D…p/M…e/C3D.any : C3D.Points.Markers.LELB.Weight.Ts : The values in Ts must be monotonically increasing

Thank you for your great support!
Best regards
Wolfgang

Hi Wolfgang,

I think the marker set should be no problem. i the mannequin file you can simply remove the expression using “RPSI” and “LPSI” position and set the numbers directly. You will have to do that for each trial maybe.

For the weight function the problem could come from the transition period. Make sure they don’t overlap each other, it could cause this kind of error.

Best regards, Sylvain.

Hi,

If u r interested in lumbar joint force, u may get wrong answers if you removed the legs in AnyBody. I am not sure about the reason behind this, but I know one big difference is psoas major muscle. You can just give it a try. If you use min/max criterion with legs, psoas major is activated, but without legs, it is not activated.

Hi Xiaopeng Jiang,

I agree with you and think that is an excellent point.

Best regards,
John

Hi Sylvain,

Two short questions:

1.What is the acceptation of Ts?
2.What do you mean with "they" in the last sentence? experimental marker?

Thank you and best regards
Wolfgang

Hi Wolfgang,

Ts is the vector giving the times when each transition occur (“The times at which transitions occur” from the reference manual).
Depending if you use the AnyFunSquareWave or AnyFunSquareWaveThreshold this vector is initialized automatically or not. Please refer to the reference manual. This sequence of time values can’t decrease. That was your error message.

The “they” was refering to transion period or transition time, the dT menber. I had the idea that if it is too long compared to the gap (time while the marker is droping) the two transition may overlap in time, or do something else that could result in that kind of error. That is if you use AnyFunSquareWaveThreshold where Ts is generated automatically.

If you use AnyFunSquareWave then i think you just typed the wrong numbers in Ts.

Best regards, Sylvain.

Hi Sylvain,

I have only one short question:):

What kind of quantity is the residual property?
I’ve read the tutorials, but i didn’t understand why the residual property gets zero during the marker dropout.

Thank you and best regards
Wolfgang

Hi

The residual value of the markers is something which comes from the motion capture, if the marker drops out the residual value will be set to zero.

Best regards
Søren

Hi Soren,

sorry for bother you again.

What would be the difference between a value of 0.5 and 0.6 for the residual property? is there a dependency on something like the distance between the marker and origin?

Thank you very much and best regards
Wolfgang

Hi Wolfgang,

The examples i have seen has been values between either zero or one.

I guess that if the system reports values in between it could mean that the marker was only partly visible by the camera’s. You will need to look into the manual of the motion capture software which produced these values to tell this exactly.

You could probably use this information to alter the weights on the markers and lower the weights on the markers with low residual values, this would technically be feasible in AnyBody, but it depends on the exact meaning of this value if it makes sense.

Best regards
Søren