Hi,
I am somewhat new to using optical tracking cameras and have been performing a variety of tests to determine how precise they are. I currently have three FLEX:V100 Optitrack cameras, which I am using with the rigid body toolkit. I intend to use the cameras to track 7/16" diameter reflective markers in a 2' x 2' x 2' space, and I wish to achieve an accuracy of under 1mm.
I recently performed a test where I fixed four markers to a rigid frame such that no three markers were co-linear and the four markers were not co-planar. I tracked these markers as I moved and rotated the frame in space. The distances and angles between each of these markers should have remained constant, yet there was a good deal of variability. I would like the variability to remain under 1-2mm and for the angle variability to remain under 5 degrees yet the variability was up to 10 times my preferred specs. I was very careful when calibrating my cameras. They are each situated 2-3 feet from the markers, and I have them positioned at various heights.
I asked a similar question a few weeks ago (http://forum.naturalpoint.com/forum/ubb ... #Post32117). Your suggestions were very helpful and made a big improvement, yet I'm in need of additional advice to address this current issue. If there is anything more I can do, I would appreciate the advice. Also, how much do you think an additional camera might help? What is the benefit of having more than three cameras?
Thanks so much for your help.
Additional Cameras?
Re: Additional Cameras?
* If it is possible, pulling the cameras back further might improve the accuracy. Their typical operating range is 4+ feet.
* A larger marker may help achieve some additional resolution.
* When operating with low camera counts, adding more cameras almost always improves the tracking redundancy and accuracy. A ray is computed between each camera and a marker, having rays from additional cameras for a marker provides more data samples for calculating the position accurately.
Getting the accuracy to be solidly under 1mm across an entire capture volume is sometimes possible but can be challenging and require experimentation with setup and wanding. You may also find that the capture volume will have a sweet spot toward its interior region.
You might also consider sending some pictures of your setup to support@naturalpoint.com . We may be able to offer additional advice based on them.
* A larger marker may help achieve some additional resolution.
* When operating with low camera counts, adding more cameras almost always improves the tracking redundancy and accuracy. A ray is computed between each camera and a marker, having rays from additional cameras for a marker provides more data samples for calculating the position accurately.
Getting the accuracy to be solidly under 1mm across an entire capture volume is sometimes possible but can be challenging and require experimentation with setup and wanding. You may also find that the capture volume will have a sweet spot toward its interior region.
You might also consider sending some pictures of your setup to support@naturalpoint.com . We may be able to offer additional advice based on them.