Connecting a wind sensor and Arduino to control a video projection

[Alexandra Gelis]

hello!
I am having a problem, see if some of you guys have any idea...
I need to trigger videos projected by tree projectors.
The videos should be trigger by the subtle air or wind produced by the walking of the viewers on a room. The viewer will  be surrounding by the 3 video projections.
My first idea was to use video cameras connected with MAXmsp and create a grid to mark and read the movement of the people on the space.
But, it will be better if  is possible to measured the air produce by their movement, because conceptually works better with the video installation that i working on it.
Any idea to how to resolve this?  
I was wondering if there is any kind of wind sensor that can be connected in to Arduino...
IDEAS?
I am really appreciated your help!

Thanks
Alexandra

[Paul Strohmeier]

you could try using piezzo electric sensors. they would trigger the projections based on the vibrations created by the peoples movements.

i think using a wind sensor will probbly be a very complex undertakig in order to get useful readings from it.

p.

- sent from android device -

--
You received this message because you are subscribed to the Google Groups "NYCResistor:Microcontrollers" group.
To post to this group, send email to nycresistormi...@googlegroups.com.
To unsubscribe from this group, send email to nycresistormicrocon...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/nycresistormicrocontrollers?hl=en.

[Eric Skiff]

Agreed - you'd probably find that any HVAC system in the space would create more noise in the wind data than people moving through the space.

Personally I would probably try to do this with a webcam and optical flow analysis using OpenCV

You can see one example of this here:

http://youtu.be/QX5TlIHXhd0

There are many other things you can do to simulate fluid/gas movement in a space once you have information about which way objects are moving in a field of view.

If you take a look at some of Grant Kot's physics simulation videos here: http://www.youtube.com/user/kotsoft?blend=8&ob=5 (he was our intern here at QLabs!) you'll get a sense for what you can do with optical flow data.

-E 

[Alexandra Gelis]

Thanks a lot!! I will try with the three options that you guys send me, thanks a lot!!!
As I will having  questions or problems I will formulating them to you, thank you very much again.
You are very kind geniuses!!
Alexandra

[Rolan]

Do ya think the Kinect would be a good fit here?

~Rolan

On Dec 13, 1:42 pm, Eric Skiff <m...@ericskiff.com> wrote:
> Agreed - you'd probably find that any HVAC system in the space would create
> more noise in the wind data than people moving through the space.
>
> Personally I would probably try to do this with a webcam and optical flow
> analysis using OpenCV
>
> You can see one example of this here:http://youtu.be/QX5TlIHXhd0
>
> There are many other things you can do to simulate fluid/gas movement in a
> space once you have information about which way objects are moving in a
> field of view.

> If you take a look at some of Grant Kot's physics simulation videos here:http://www.youtube.com/user/kotsoft?blend=8&ob=5(he was our intern here at

> QLabs!) you'll get a sense for what you can do with optical flow data.
>
> -E
>
> On Tue, Dec 13, 2011 at 1:03 PM, Paul Strohmeier

> <paul.strohme...@gmail.com>wrote:

>
>
>
>
>
>
>
> > you could try using piezzo electric sensors. they would trigger the
> > projections based on the vibrations created by the peoples movements.
>
> > i think using a wind sensor will probbly be a very complex undertakig in
> > order to get useful readings from it.
>
> > p.
>
> > - sent from android device -

> > On Dec 13, 2011 6:02 AM, "Alexandra Gelis" <i...@alexandragelis.com>

[michael lowry]

Kinect sounds like fun in this situation. Also, you could opt for IR or Sonar rangefinders on a grid or in some sort of arrangement around the room as well. 

Michael Lowry
Research Assistant

NYU Minimally Invasive Surgery Lab

423 E 23rd Street, New York, NY 10010

Office: (212) 686-7500 X6441
Cell: (201) 787-8633

mt...@njit.edu

[Dave Pentecost]

Check Memo Akten's MSAFluid for Processing and Open Frameworks, for a
fluid simulation that could be manipulated by CV and in turn
manipulate the sound.

http://memo.tv/msafluid_for_processing_v1_3
http://memo.tv/ofxmsafluid

Dave

--
Director, Technology
Center for Community
http://www.girlsclub.org/building

Community Base
http://c4c.posterous.com

@dpentecost
Cell 646 704 2021

[Dan Lavin]

Have you considered a passive infrared (PIR) sensor? These are commonly
used in motion detectors. One or more of these may do what you want.
> http://www.parallax.com/Store/Sensors/ObjectDetection/tabid/176/CategoryID/51/List/0/SortField/0/Level/a/ProductID/83/Default.aspx

Alternatively, you could try hanging some sheets of aluminum foil in the
room and detect their movement with a reflective sensor.

[Alexandra Gelis]

I really like the idea of using Kinetics.. do you more info about it? some tutorial? or suggestions that I can follow?
Thanks a lot!!
Alexandra

nycresistormicrocontrollers@googlegroups.com.

To unsubscribe from this group, send email to

nycresistormicrocontrollers+unsub...@googlegroups.com.

For more options, visit this group at
http://groups.google.com/group/nycresistormicrocontrollers?hl=en.

 --
You received this message because you are subscribed to the Google Groups
"NYCResistor:Microcontrollers" group.
To post to this group, send email to

nycresistormicrocontrollers@googlegroups.com.

To unsubscribe from this group, send email to

nycresistormicrocontrollers+unsub...@googlegroups.com.

For more options, visit this group at
http://groups.google.com/group/nycresistormicrocontrollers?hl=en.

--
You received this message because you are subscribed to the Google Groups "NYCResistor:Microcontrollers" group.

To post to this group, send email to nycresistormicrocontrollers@googlegroups.com.
To unsubscribe from this group, send email to nycresistormicrocontrollers+unsub...@googlegroups.com.

[P Leiby]

Here are Windows-based resources for using Kinect:

Coding for Fun Kinect projects: http://channel9.msdn.com/coding4fun/kinect/

"Kinect for Windows SDK Quickstarts" (http://channel9.msdn.com/series/KinectSDKQuickstarts/) 

- (5 or 6 videos on setting up Kinect and Dev environment, Camera Fundamentals, Depth Fundamentals, Skeleton Data, Audio Fundamentals)

To post to this group, send email to nycresistormi...@googlegroups.com.
To unsubscribe from this group, send email to nycresistormicrocon...@googlegroups.com.

[Eric Skiff]

We've used the Kinect successfully to do more precise things where CV would normally be a good fit, because with the depth data we can easily filter out the background.

-E

On Tue, Dec 13, 2011 at 3:50 PM, Rolan <rola...@gmail.com> wrote:

[Alexandra Gelis]

So Eric your suggestion it will be work with CV? do you have some examples, or tutorials?
I haven work before with CV at all....
Thanks!
Alexandra

[Eric Skiff]

The openFrameworks OpenCV demos are a fantastic starting point! Are you on PC or Mac?

http://www.openframeworks.cc/download

Presuming you're on a mac and have xcode, you'd download the latest openframeworks, go into the addons folder, then openCV, and open one of the demo projects in there. The vast majority of openFrameworks projects I've seen start from one of the demos and then tweak/build from there.

The ofxKinect package is very similar, once you've got the basics of working with openFrameworks down, it just provides depth data along with the pixels, which can be much cleaner to work with if you're using it to affect an interactive installation.

-E