HELP:PRECISION OF V100:R2 CAMERAS!?!

[la_luna]

Dear,

My current work is on human gait analysis for what I use OptiTrack system.

I computed standard deviation for markers position in x,y and z coordinate. Standard deviation for x and z coordinate was in acceptable limits, but standard deviation for y coordinate was unacceptable. For instance, probationer walks in length from 1,5m to 3,0m along y direction and my standard deviation for that length is from -50,0mm to 50,0mm or more.

My question is what is precision of V100:R2 camera?

If I know that value, maybe I'll find mistake that I made.

Best regards

[NaturalPoint - Mike]

Before I can answer this, I need to know a bit more information about your setup.

How many cameras do you have, and how are they arranged?
What software are you using to run the cameras (Arena, Tracking Tools, Tracking Tools API, Camera SDK)?
How are the markers arranged? Are you defining rigid bodies?

[la_luna]

Dear,

I have 6 cameras and they're arranged along walk direction (e.g. y-axis). For software, I'm using Arena.

I'm using 4 markers for 4 characteristics points on human leg which are placed in lateral side of leg (e.g. they are in the sagittal (profile) plane).

Off course, that I define rigid body first, and after that I'm recording the movement of probationer.

Best regards

[NaturalPoint - Mike]

Do you have two rigid bodies defined or are you defining them as one?

Properly calibrated systems should have a margin of error in marker position reporting of less than 1mm.

Could you perhaps send some .pt2 and corresponding .rgb files to support@naturalpoint.com for review?

[la_luna]

Dear Mike,

Thank you for your quick answer. Download link for .pt2 and .rgb files is https://rapidshare.com/files/2759001409/Results.rar.

Best regards

[NaturalPoint - Mike]

File has been downloaded. I'll let you know what I see here.

[la_luna]

Thank you very much. I'm that your's comments will help me.

[NaturalPoint - Mike]

Hmm, looking at the data I see a few things. First and foremost, the aiming is a bit high, and that's what's causing the lower markers to drop out. You may want to move the cameras back a bit and aim them so they can see upper and lower properly Positioning them so that they have a bit of variance in height would likely help as well.

Additionally, I notice alot of masking in the camera views. Is this reflection off of something? It's always best to use masking as a last resort, and try to find a way to physically block or remove the source of the IR interference.

[la_luna]

Dear Mike,

Thank you for your comments.

Well, that reflection that you mentioned was from light, and I can fix that.

Guided by your comments, I hope that I'll be able to finish this testing in the right direction.

Best regrads,

[NaturalPoint - Mike]

Not a problem. If you'd like me to take a look at the new data once you've made a change, let me know!