Page 1 of 1

Cluster Markers' function in motion analysis

Posted: Thu Jan 08, 2015 10:10 am
by shuyanghan
Hi everyone,

I have a basic question for motion analysis here. In the Cleveland Clinic marker set, there are some cluster markers (such as L.Thigh.Upper, L.Thigh.Front, etc.). What is the specific function of these makers? Because it seems that the segment could also be well defined without these markers (for example, definition of the knee joint by Knee.Lat, Knee.Med, Ankle.Lat, and Ankle.Med). I suppose these markers help increase the accuracy of results, but how?

Here I also have another question. I am a new user of OpenSim. In my project, I only need to capture the lower limb activities. In OpenSim, I use the 2392 model, and delete the markers for upper limb, and those on the foot except for Heel and Toe.Tip. While scaling the model, in Tab "Scale Factors", the row of torso, talus becomes red, there is a ? in the last column. Should I manually fill in these numbers? And then, when doing IK, the number of Weight are all 1.0. And I encountered the following problem showing in the message window:
InverseKinematicsTool Failed: SimTK Exception thrown at Assembler.h:1321:
Error detected by Simbody method Markers::moveAllObservations(): Number of observations provided (8) differs from the number of observations (14) last defined with defineObservationOrder().
(Required condition '(int)observations.size() == (int)observation2marker.size()' was not met.)



Thank you !

Re: Cluster Markers' function in motion analysis

Posted: Thu Jan 08, 2015 10:28 am
by shuyanghan
One more thing to mention is that the .trc file was generated within Cortex, we are using Motion Analysis Corp.

Re: Cluster Markers' function in motion analysis

Posted: Mon Jan 12, 2015 9:30 am
by shuyanghan
Hello Dear OpenSim community,

I have solved the second problem after practicing with the tutorial and user guide. In terms of the first question, does anyone know the function of the markers attached to the thigh and shank ? I would be very delighted if anyone could post some suggestions. Thank you.

Re: Cluster Markers' function in motion analysis

Posted: Wed Jan 14, 2015 12:29 pm
by jimmy
Hi Shuyang-

The more markers you have in the minimization equation, the less the coordinates are effected by error at any single point. If you had only one or two markers on a segment, and there was alot of local noise, tracking would be highly effected.

Every marker on skin is effecedt by soft-tissue artifact. Markers on joint axes can be particularly influenced by local skin stretch changes so segment tracking tends to be preferred during IK. To reduce the effect of any single noisey marker, clusters are used.

Cheers,
-james

Re: Cluster Markers' function in motion analysis

Posted: Sun May 24, 2015 6:32 pm
by mohsensharifi
shuyanghan wrote:Hi everyone,

I have a basic question for motion analysis here. In the Cleveland Clinic marker set, there are some cluster markers (such as L.Thigh.Upper, L.Thigh.Front, etc.). What is the specific function of these makers? Because it seems that the segment could also be well defined without these markers (for example, definition of the knee joint by Knee.Lat, Knee.Med, Ankle.Lat, and Ankle.Med). I suppose these markers help increase the accuracy of results, but how?

Here I also have another question. I am a new user of OpenSim. In my project, I only need to capture the lower limb activities. In OpenSim, I use the 2392 model, and delete the markers for upper limb, and those on the foot except for Heel and Toe.Tip. While scaling the model, in Tab "Scale Factors", the row of torso, talus becomes red, there is a ? in the last column. Should I manually fill in these numbers? And then, when doing IK, the number of Weight are all 1.0. And I encountered the following problem showing in the message window:
InverseKinematicsTool Failed: SimTK Exception thrown at Assembler.h:1321:
Error detected by Simbody method Markers::moveAllObservations(): Number of observations provided (8) differs from the number of observations (14) last defined with defineObservationOrder().
(Required condition '(int)observations.size() == (int)observation2marker.size()' was not met.)



Thank you !
Hi shuyanghan and all,

I am a new user of Opensim and trying to use our experiment data in Opensim. But I have encountered a problem like the second your problem. I was wondering how you fixed it.

Here is the error:
Error detected by Simbody method Markers::moveAllObservations(): Number of observations provided (14) differs from the number of observations (15) last defined with defineObservationOrder().
(Required condition '(int)observations.size() == (int)observation2marker.size()' was not met.)

I would be thank full if any one else could help me in this case.

Thanks,
Mohsen