V2:Trio refine calibration

Post Reply
vgarciavazquez
Posts: 3
Joined: Fri Dec 24, 2021 2:42 am

V2:Trio refine calibration

Post by vgarciavazquez »

Hi,

V120:Trio is self-calibrated but is there a way to refine that calibration? We are using that system and Motive 2.3.0 Final to find the transformation (specifically a translation in x, y and z) between two rigid bodies attached to a rigid board. The transformation changes (for example, norm of the difference of two transformations was 8 mm) when changing the position of that board although that transformation should be more or less constant. One idea could be that the cameras are miscalibrated so we could refine the calibration with Motive but, if I am not wrong, Motive does not show the Camera Calibration Panel when using V120:Trio.

I have seen this post (not acceptable tracking performance with V120:Trio) but they used Tracking Tools version 2.3.2 (build number 10268)

viewtopic.php?f=56&t=9568&p=48300&hilit ... acy#p48300

Has anyone checked the accuracy of V120:Trio?

Just in case, our tracking bar serial number is 149816.

Thanks!
vgarciavazquez
Posts: 3
Joined: Fri Dec 24, 2021 2:42 am

Re: V2:Trio refine calibration

Post by vgarciavazquez »

The residual error was around 0.7 mm.

https://v23.wiki.optitrack.com/index.ph ... alibration explains how to update the calibration of Duo/Trio Tracking Bars to improve tracking quality. With this feature the residual error is around 0.08 mm and the transformations are more constant now (mean difference of 1.2 mm) although we obtained a maximum difference between two transformations of around 4.5 mm.

We have set the following settings to improve accuracy and stability:

MaxMarkerDeflection = 0.002 mm
MinimumMarkerCount = 4 (both rigid bodies have 4 markers)
AcquisitionMarkerCount = 4
Unique = TRUE
Flexibility = 0.1
TrackingAlgorithmLevel = RigidBodySolver::cRigidBodySettings::Algorithm_RayBasedUltimate
ShareMarkers = FALSE
JumpPrevention = TRUE
LocalizeSearch = TRUE
Maximum Residual: 0.8 mm
Minimum Ray Count: 3
Circularity: 0.8
Exposure: 20 us
Threshold: 240

V120:Trio placed at 1.2 m from both rigid bodies.

Any suggestion to improve those results? Has anyone checked the accuracy of V120:Trio?

Thanks
Post Reply