Change font size
It is currently Thu Feb 21, 2019 1:48 am


Post a new topicPost a reply Page 1 of 1   [ 8 posts ]
Author Message
PostPosted: Sat Jun 02, 2012 12:52 pm 

Joined: Sun Apr 12, 2009 8:38 pm
Posts: 16
Hi there,

I'm planning on setting up a 2 Kinect capture area in a small spare room that is 9-foot by 9-foot.

What would be the optimal placement of the cameras for best capture? I was thinking of placing
1 Kinect in a corner at about 5-foot above the floor and the other Kinect in the other opposite
corner at about 2-foot off the floor. Hoping to mount them somewhat 'permanently', so want to
do it right the first time. ;)

Can anyone share their experience/advice on this?

Thanks! :)


Top
 Profile  
 
PostPosted: Sat Jun 02, 2012 2:46 pm 

Joined: Mon Aug 03, 2009 1:34 pm
Posts: 2320
Location: Los Angeles
9 x 9 is a little tight space for mocap--it might be just enough if you perform in a corner but you probably won't be able to take more than a step or two in any direction.

I think I would try placing the devices near the middle of two adjacent walls, angle the devices towards the far quadrant/corner of the room and do my performance in that quadrant. This will give you about 90 degree recording and probably the largest capture space. Normally I position my Kinects maybe 4-5 ft off the ground, but if you set it higher you may actually get more coverage. I would start with this setup but be sure to experiment and see what works best for you.

Alternatively, you might place Kinects in far corners for near 180 capture, but I don't think the room will be large enough for this. Also, you may get more IR errors because the devices will be so close to each other. IMO, the 90 degree option may be more practical for a small room but try it and see what happens.

BTW, I would advise against permanently fixing the devices to the walls as, depending on the motion you're trying to record, you may want to raise or lower them at times to improve coverage. But if you're determined to do so, make sure you do plenty of recording and tracking first to see which position gives you the best results.

Are there large windows in the room? If so, you may want to invest in blinds or curtains. The reflective surface of the glass and bright outdoor light can cause errors.

Good luck!

G.

_________________
Greenlaw
Artist/Partner - Little Green Dog | Demo Reel (2017) | Demo Reel (2015) | Demo Reel (2013)

Image
Watch a one minute excerpt on Vimeo now!


Top
 Profile  
 
PostPosted: Sat Jun 02, 2012 6:34 pm 

Joined: Sun Apr 12, 2009 8:38 pm
Posts: 16
Thanks for the feedback, Greenlaw (was hoping you'd chime in) ;)

So of the two options below, #2 would probably result in a better
quality capture? (My rationale for putting a Kinect at 2(ish)-feet off
the floor was to hopefully pick up any lower areas the other Kinect
may have had trouble catching...but this was just theory)

Image
Image

Hopefully this thread will assist other folks who may be attempting to capture
in such small area...


Top
 Profile  
 
PostPosted: Sat Jun 02, 2012 8:19 pm 

Joined: Mon Aug 03, 2009 1:34 pm
Posts: 2320
Location: Los Angeles
This is what I was picturing in my head, which is a variation of configuration number 2:
Attachment:
Configuration3.jpg
Configuration3.jpg [ 13.84 KiB | Viewed 7105 times ]

I'm not sure if this is any better than 2 but there you go. I thought this configuration might provide a deeper capture space but it looks like number 2 might give you a wider space.

Let us know how things work out for you.

G.

_________________
Greenlaw
Artist/Partner - Little Green Dog | Demo Reel (2017) | Demo Reel (2015) | Demo Reel (2013)

Image
Watch a one minute excerpt on Vimeo now!


Top
 Profile  
 
PostPosted: Sun Jun 03, 2012 12:58 am 

Joined: Mon Aug 03, 2009 1:34 pm
Posts: 2320
Location: Los Angeles
I adjusted the positions a bit and this seems even bigger plus you get full the 90 degrees. I'm not sure how well it translates into real world use but give it a try.


Attachments:
File comment: Configuration #3 Adjusted
Configuration3_Adjusted.jpg
Configuration3_Adjusted.jpg [ 7.21 KiB | Viewed 7098 times ]

_________________
Greenlaw
Artist/Partner - Little Green Dog | Demo Reel (2017) | Demo Reel (2015) | Demo Reel (2013)

Image
Watch a one minute excerpt on Vimeo now!
Top
 Profile  
 
PostPosted: Sun Jun 03, 2012 1:11 am 

Joined: Mon Aug 03, 2009 1:34 pm
Posts: 2320
Location: Los Angeles
Regarding low position of one Kinect: having staggered camera heights seemed to help with PS3 Eye cameras, but I'm not sure if this helps much with Kinect. It may help depending on the motion you're trying to capture but in general I think you'd be better off keeping them closer to chest-to-eye level.

FYI, I try to keep the camera views more or less mirrored. To do this, I place an object on the floor in the center of the capture space and align the devices/views so that the object is in relatively the same place for each view. A human volunteer is best because you can adjust for actual height but I've used a chair and a even a book on the floor. For me, the remote controlled motor in the Kinect is useful for adjusting the alignment this because my computer is located in a different room from where I shoot. Once the cameras are aligned, I mark the position on the floor of the object/person with a piece of gaffers tape and remove the object. This is where the performer does his T-Pose.

G.

_________________
Greenlaw
Artist/Partner - Little Green Dog | Demo Reel (2017) | Demo Reel (2015) | Demo Reel (2013)

Image
Watch a one minute excerpt on Vimeo now!


Top
 Profile  
 
PostPosted: Sun Jun 03, 2012 10:23 am 

Joined: Sun Apr 12, 2009 8:38 pm
Posts: 16
awesome! Thanks for the follow-up thoughts, Greenlaw.

When I get set-up I'll post how it turned out. :)


Top
 Profile  
 
PostPosted: Sun Jun 03, 2012 11:03 pm 

Joined: Mon Aug 03, 2009 1:34 pm
Posts: 2320
Location: Los Angeles
On second thought, staggering the heights, at least slightly, may reduce IR interference. This is just a guess based on past observation, and it's probably easy enough to test.

If you see a lot of yellow colored pixels in the Kinect depth data, those are unreadable surfaces, and you want to minimize that. I believe these 'yellow' zones are caused by fully black surfaces, extremely bright light, highly reflective surface, and IR bounces coming from the other Kinect. Changing the environment and the position/angle of the Kinects should visibly change the amount of 'yellow' data that appears on the screen.

It's probably more important to minimize 'yellow' data on the performer's volume but you don't want to see a lot in the background either.

Hope this helps.

G.

_________________
Greenlaw
Artist/Partner - Little Green Dog | Demo Reel (2017) | Demo Reel (2015) | Demo Reel (2013)

Image
Watch a one minute excerpt on Vimeo now!


Top
 Profile  
 
Display posts from previous:  Sort by  
Post a new topicPost a reply Page 1 of 1   [ 8 posts ]


Who is online

Users browsing this forum: No registered users and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  


Powered by phpBB® Forum Software © phpBB Group
610nm Style by Daniel St. Jules of Gamexe.net