Hi!

on which platform are you working? TuioClient is using libfreenect to access kinect. so it won't work simultaneously with pix_freenect or fux_kinect. i don't know if the linux kernel drivers work in parallel to libfreenect but i don't think so.

with pix_freenect (1) you can grab rgb and depth streams from kinect sensor. you can do the (hand)tracking in pd by using pix_opencv tools and/or pix_multiblob. also pix_curve can be useful to get images of certain distance out of the depth stream from the kinect. (in combination with switching into yuv colorspace)

if you decide for pix_openni (2) you can use the hand tracking or skeleton tracking from nite/openni together with the image and depth stream.

matthias

(1) http://github.com/kronihias/pix_freenect (2) http://github.com/kronihias/pix_openni

Am 22.01.12 23:22, schrieb Philipp Wartenberg:

Hello,

is it possible to use the video stream of the kinect simutaneously with [pix_video] and [TuioClient]?

I want to be able to define separated areas as a layer over the video stream to get seprated tracking areas.

For example: If movement is detected in a specified area the x will happen...

The difinition of areas needs to be changed several times and thats why I want to be able to "sketch" over the video picture. If you know the EyeCon System, you may know what I mean.

Kind regards, Philipp

---

Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list