[Fluxus] kinetic

[Scott]

We're thinking about this at work too... 1) Obviously, one could just
use the raw data - a color picture and associated depth map, or 2) one
could use the XBOX SDK for this and output the raw data AND skeletal
information, i.e. limb positions and joint angles. The you just attach
a whole xbox as an IO device.

Doing #1 is straight forward, and should just be done since no xbox
required, #2 provides a lot of "pre-digested" information that can
make writing apps easy.

On Tue, Nov 16, 2010 at 1:23 PM, Kassen <signal.automatique at gmail.com> wrote:
>>>
>> it would be a nice fluxus module indeed. i think it would need its own
>> special module, since the kinect also provides the depth map and other
>> useful information apart from the video image. we were planning a plugin
>> system with dave that would allow us to interface modules with internal
>> fluxus primitive types and data easily. this way the module could send the
>> video and the depth map as two pixel primitives.
>>
>
> Great stuff!
> That demo I linked to makes it look a bit like it acts as a webcam that
> creates a single image consisting of two "viewports" side by side, with the
> left one being the depth-map. A image like that could be treated by Fluxus,
> but I shudder to think of the amount of per-pixel-per-frame processing.
> At any rate the future looks bright, thanks to the affordable and open tools
> of our friends in Redmond! ;¬)
> Yours,
> Kas.