Questions regarding Rigid Bodies

Post Reply
sceu
Posts: 2
Joined: Mon Aug 28, 2017 2:00 am

Questions regarding Rigid Bodies

Post by sceu »

hi,

I am quite new to Motive and Optitrack. I did some test measurements of devices and humans to become familiar with the system. Currently I am working with Motive 1.10.

Now I have a couple of questions regarding rigid bodies, hopefully someone can help me :-).
Many thanks in advance!


Question 1: Benefits of rigid bodies for marker tracking

If I create a rigid body before recording, the defined rigid body can improve the marker tracking ("Reconstructed Markers Only" = false, "Tracking Algorithm" = Ray based), right?
If I create a rigid body in an existing take, it will not have any positiv effect on marker tracking anymore?

Question 2: Renaming markers in rigid bodies

Is it possible to rename the markers of a rigid body in Motive 2.0?

Question 3: Experiments with rigid bodies (labeling of markers)
I created a rigid body with the name Arm3 consisting of 4 markers in an existing take. The take was recorded without rigid bodies or a marker set. As I wanted to define the positions of the markers within the rigid body, I manually labeled the markers. This was the result:
Image
So the position of each marker within a rigid body is predefined by Motive? Is there a way to change/ define the positions/ oder of the marker in a rigid body manually?

If I apply "Reconstruct and Auto-label" on the take with the described rigid body Arm3, the orientation of the rigid body based on the 4 markers is correct. However, the rigid body "jumps away" at a certain point in the take but moves back into the correct position a few frames later:
Image
What could be the reason for the this "jump"?

If I manually label the markers of the rigid body Arm3 in a slightly different order compared to the auto-labeling, the body does not jump anymore:
Image
How does Motive determine the position of the markers within a rigid body? And how can I make sure, a rigid body does not jump?

Question 4: Define axis of rigid bodies
I am interested in the orientatation of rigid bodies and the angle between two rigid bodies. For this, it would be useful to manually define the rotation axis of rigid bodies. It is possible to set the coordinate system of a rigid body according to the global coordinate system. Unfortunately, this is too inaccurate, I would like to define the coordinate system based on the rigid body makers for example. Do you have a hint for me how to define rotation axes or how to calulate angles/ orientation of rigid bodies?
In the meantime I solved the "problem" by importing the coordinates of the markes as .csv into Matlab and calculate orientation/ angles of the desired segments.

Last Question 5: Calibration/ Re-calibration
If I reopen a calibrated project without changing anything on the hardware setup, do I need to perform a re-calibration and re-definition of the ground plane?


Sincerely,
sceu
steven.andrews
NaturalPoint Employee
NaturalPoint Employee
Posts: 737
Joined: Mon Jan 19, 2015 11:52 am

Re: Questions regarding Rigid Bodies

Post by steven.andrews »

Hello sceu,

Thank you for reaching out to the community regarding your Rigid Body questions. I have some answers for you, below.

Question 1: Benefits of rigid bodies for marker tracking
If I create a rigid body before recording, the defined rigid body can improve the marker tracking ("Reconstructed Markers Only" = false, "Tracking Algorithm" = Ray based), right?
If I create a rigid body in an existing take, it will not have any positiv effect on marker tracking anymore?

If you create a Rigid Body in Edit Mode and use the above settings, those settings should still affect the tracking results. After creating the Rigid Body, you will typically need to Reconstruct and Autolabel for the Rigid Body to be utilized.

Question 2: Renaming markers in rigid bodies
Is it possible to rename the markers of a rigid body in Motive 2.0?

This is not currently possible for a Rigid Body, but marker names and colors can be changed for a Skeleton or a Markerset. We do plan to add this capability for Rigid Bodies.

Question 3: Experiments with rigid bodies (labeling of markers)
I did some testing manually labeling Rigid Bodies in Motive 2.0, and this appears to be working in the current version. If you continue to encounter issues with this in 2.0, please contact us directly so we can troubleshoot with you.

Question 4: Define axis of rigid bodies
I am interested in the orientatation of rigid bodies and the angle between two rigid bodies. For this, it would be useful to manually define the rotation axis of rigid bodies. It is possible to set the coordinate system of a rigid body according to the global coordinate system. Unfortunately, this is too inaccurate, I would like to define the coordinate system based on the rigid body makers for example. Do you have a hint for me how to define rotation axes or how to calulate angles/ orientation of rigid bodies?

When first created, the orientation of a Rigid Body will exactly match the orientation of the global coordinate system. From here, it is possible to apply translations and rotations to the Pivot Point in order to change how the rigid body solves within its marker cloud. These links will take you to the documentation for doing this in Motive 2.0, where the widgets have been improved for this kind of manipulation. You can manually enter values to change the orientation of the Rigid Body, or you can use the widgets to align the Pivot Point to the markers.
http://v20.wiki.optitrack.com/index.php ... ivot_Point
http://v20.wiki.optitrack.com/index.php ... _and_Scale

Last Question 5: Calibration/ Re-calibration
If I reopen a calibrated project without changing anything on the hardware setup, do I need to perform a re-calibration and re-definition of the ground plane?

In general, we recommend recalibrating your system regularly. If cameras are mounted securely, it is possible a calibration can last for days or longer. But this is dependent on the mounting and the environment. Slight vibrations on the cameras, mounting surfaces and even temperature change can throw off a calibration. My rule of thumb is calibrating before every new session.
If you do not want to reset the ground plane, you can perform a Refine calibration, which will help to fine-tune your already calibrated cameras without resetting the ground plane.

Motive 2.0 contains a Continuous Calibration feature that helps to increase the fidelity of your calibration over a longer period of time. If you are able to update to this version, the feature would help with this.



I hope you find all of the above to be useful. If you require any further assistance, please feel free to reach out to our Support team directly.

Best,
Steven
--
Steven Andrews
OptiTrack | Senior Customer Support Engineer
help.naturalpoint.com
support@optitrack.com
Post Reply