Kinect v2 and the Vive

[Remi Rousseau]

Hey guys,

I'm working on a project where a user's body is captured with two Kinect v2 and displayed in pointclouds on the Vive.

The big concern is that there are big interferences between the Lighthouse and the Kinect V2 that lead to a big drift or a loss of tracking for both the controllers and the headset... Not a pleasant experience!

Do you have any idea on solutions to try in order to bypass this issue? We tried with several filters on the Kinect, with very few enhancement

Thanks a lot!

Rémi

[Laurent BOIREAU]

I'm afraid you're stuck here: both systems use structured light, and I'll wager similar (infrared) wavelength since they interfere. I would try to place the kinect orthogonal to the lighthouses orientations as a possible mitigation, but that is a (very) long shot.
You may consider however the use of an alternative system: Intel Realsense depth cameras, for instance, use Time Of Flight rather than pattern projection. Next time I get near a Lighthouse, I'll bring one to check if these would be immune to interference, and let you know. Hopefully this week !

They have better depth resolution than Kinect, albeit shorter range (2.5 or 3m).

          Laurent

--
You received this message because you are subscribed to the Google Groups "VR Geeks" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vr-geeks+u...@googlegroups.com.
Visit this group at http://groups.google.com/group/vr-geeks.
For more options, visit https://groups.google.com/d/optout.

[Christophe Gouet]

Hi Rémi and VRGeeks !

Just a quick idea: would cheap passive circular polarizing filters do the job?

I mean one channel on the Kinect V2 and the other channel on the Vive's Lighthouses.

Best regards,

Christophe

[Lorne Covington]

Neither the K2 nor the Vive use structured light; the K2 is time-of-flight, and the Vive lighthouses appear to use swept X/Y lines and a reference pulse emitter, with position determined by timing the X/Y line crossings relative to the reference pulse.  The interference is likely the K2 light pulses being mistaken for one of those signals.

So the Intel cameras will likely cause the same problem.  The IR filters may help, but the Vive lighthouses likely use IR lasers on the scanners, not LEDs, and those may be polarized - meaning the filters could reduce their intensity.  Worth trying though, I have some filters, and was going to try the Vive with a K2 and you just saved me a step!

- Lorne


http://noirflux.com

[Remi Rousseau]

We indeed tried with the realsense and had also interferences, though less annoying than with the kinect  2 (which is more powerful)

No interferences on the kinect 1 by the way

--

You received this message because you are subscribed to a topic in the Google Groups "VR Geeks" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/vr-geeks/4xM1qlzoWlE/unsubscribe.
To unsubscribe from this group and all its topics, send an email to vr-geeks+u...@googlegroups.com.

[Lorne Covington]

Good to know.  Makes sense about the K1, since it is structured light and not pulsed.  That means the other SL cameras like the Primesense (RIP), Xtions, and the new Orbec also may not interfere.

- Lorne


http://noirflux.com

[Jan Ciger]

On Tue, Dec 1, 2015 at 2:26 PM, Lorne Covington <noir...@gmail.com> wrote:
>
> Good to know.  Makes sense about the K1, since it is structured light and
> not pulsed.  That means the other SL cameras like the Primesense (RIP),
> Xtions, and the new Orbec also may not interfere.

Is Xtion actually still available? That thing uses the same sensor as original Kinect and Primesense (the Primesense chipset). And those aren't available anymore after they have been bought out by Apple, or?

The Vive uses IR lasers, indeed.

On 12/1/2015 8:06 AM, Remi Rousseau wrote:
We indeed tried with the realsense and had also interferences, though less annoying than with the kinect  2 (which is more powerful)
No interferences on the kinect 1 by the way

I am afraid that the only solution is to simplify the system - use only the Kinects for position tracking and not the Lighthouse system. The Kinect position tracking data for the head and hands are fairly OK from my experience, the Kinect 1 was quite jittery but Kinect 2 is good enough even without extra filtering.

Alternatively, switch to a non-optical tracking system for the HMD, if the Kinect data are not good enough (e.g. because of range).

IR light interference is an unfortunate fact of life. IR LEDs commonly come in two versions - 850nm and 940-960nm. Most of the tracking systems use near IR, around 850nm (you can see it as a dark red glow) because that is where the camera sensors used are still sensitive. So they will interfere with each other.

The other common IR wavelength is 940-960nm, that is not visible to human eye anymore. Unfortunately, it is poorly visible for CCDs as well, so it is rarely used for tracking, mostly for things like TV remotes and such.

Filters are not of much use, because all the systems are on the same wavelength. Polarization could be tried, but it likely wouldn't work with lasers - laser light is often polarized already due to the way it is generated.

Commercial systems do the same thing - e.g. our Optitrack at work reliably interferes with any IR synchronized shutter glasses because of the flashing and intensity of the IR illuminators drowning out the sync signal.

Jan