I have a friend who got some for biology research. Apparently they compress the total depth that the Kinect is seeing, but it still can't "focus" up close.
I'll buy a pair, it's not much money to try something interesting.
Microsoft Kinect: infrared depth maps for dewarping?
Moderator: peterZ
- daniel_reetz
- Posts: 2812
- Joined: 03 Jun 2009, 13:56
- E-book readers owned: Used to have a PRS-500
- Number of books owned: 600
- Country: United States
- Contact:
Re: Microsoft Kinect: infrared depth maps for dewarping?
Unsure if this deserves a separate post.... or not as it is not about Kinect.... but another option that may move this ahead faster (and cheaper)...
Point Grey makes top notch cameras with full access to drivers that may be interesting....
http://www.ptgrey.com/products/fireflym ... camera.asp
Now they are usually expensive... but there is a cache of them available at a local recycling place that I frequent.... A certain international touch interface company (a very intelligent one ) was using these cameras in their multi-touch tables. They retail for about $300 ($420 including IR filter, lens mount and lens). Well... said company changed specs... and had about 2,000 of them to unload..... I am buying them at $25 each (new in box) including an IR filter, lens mount, and lens. They shoot at 60 FPS! I am about 2/3 of the way through building a multi-touch table for my kids.
There is another active and supportive maker community that specializes in building multitouch tables/surfaces http://nuigroup.com/forums. They have opensource SW (Community Core Vision) for managing the camera and the multiple input points (IR points of light detected by the cameras). The SW allows you to easily receive/record the data about which points are being detected.
You can also get parts (Lasers, mounts, line filters, power adapters, etc.) from Peau Productions -http://www.peauproductions.com/main.html
My thinking is that if you just use laser dots (instead of grids) you may be able to then use the points detected from the opensource CCV product to then produce the page specific algorithm to remove curvature of the page. Maybe this was already covered previously in this or another forum and I am just late to the party (wouldn't be the first time).
Cheers
Point Grey makes top notch cameras with full access to drivers that may be interesting....
http://www.ptgrey.com/products/fireflym ... camera.asp
Now they are usually expensive... but there is a cache of them available at a local recycling place that I frequent.... A certain international touch interface company (a very intelligent one ) was using these cameras in their multi-touch tables. They retail for about $300 ($420 including IR filter, lens mount and lens). Well... said company changed specs... and had about 2,000 of them to unload..... I am buying them at $25 each (new in box) including an IR filter, lens mount, and lens. They shoot at 60 FPS! I am about 2/3 of the way through building a multi-touch table for my kids.
There is another active and supportive maker community that specializes in building multitouch tables/surfaces http://nuigroup.com/forums. They have opensource SW (Community Core Vision) for managing the camera and the multiple input points (IR points of light detected by the cameras). The SW allows you to easily receive/record the data about which points are being detected.
You can also get parts (Lasers, mounts, line filters, power adapters, etc.) from Peau Productions -http://www.peauproductions.com/main.html
My thinking is that if you just use laser dots (instead of grids) you may be able to then use the points detected from the opensource CCV product to then produce the page specific algorithm to remove curvature of the page. Maybe this was already covered previously in this or another forum and I am just late to the party (wouldn't be the first time).
Cheers