Say Cheese!

I've been testing out the Canon SDK for an powershot model I have and I finally got it to work! I made a simple console app that starts the sdk, connects to the camera, and then takes a pictures and saves it to the computer. Not bad for a few hours of shuffling through documentation ;) How this works into my multitouch project... that's my secret for now!

In other news, I made the final screen for my multitouch box using the glue method I previously described. I'll admit, it was much more challenging than I initially thought, simply because of the size. I ended up diluting the glue to a thin consistency and then spraying it on the acrylic. I then used a large squeegee to get all the excess glue and air out from under the rosco grey. Overall, the screen looks great. I should be able to get it reinstalled into the big box so I can test the image quality.

Photo App Video!

So, I've been working on the basecode for my multitouch setup for a while now. I made this simple photo app as a test. It still needs a little work, but I'm pretty pleased with it so far. Take a look!

Smoke and Mirrors

Well I got my projector today! I took over the spare bedroom in the house and setup a makeshift table to test things out. I wanted to check for a few things...

a) how difficult would it be to get the projector image straight when going from horizontal to vertical via the mirror reflection

b) how much infrared light the projector gave off

c) if the projector would look washed out with some regular lights on

Here's what I found out. It was pretty easy to get the projector to show up on my screen. I was even more surprised with how well vellum worked as a projection surface. When I turned on my IR webcam to check the projector it picked up virtually no IR light. It gets even better though, because even with with all the lights on in the room the projection was still crisp and the contrast looked great. All in all... I'm feeling pretty good about the project. Check out some pics!

More waiting

Well after my last post I wanted to try illuminating my surface from behind with a few IR LEDs. I found that with everything I tried, the LEDs just put way too much focused light on the surface. After talking with cerupcat on the nuigroup forums, he mentioned the use of the GamFusion material from www.gamonline.com as a diffuser of the IR light. The stuff is pretty cheap so I ordered some along with some more LEDs.

On a more exciting note, I came into some unexpected funding for my project. I can finally afford to buy a projector :)

Change of Direction

Well a few days ago I recieved my silicone-rubber sheet from Gravano of nuigroup.com. After getting it setup on my table and doing some more work on my software I was able to get pretty good finger tracking. The problems came the next day, when my software went crazy because of a little bit of ambient daylight in the room. Since I want to use this technology someday in the classroom, this presents a roadblock.

So, I've now decided to shift my focus from FTIR (sides of acrylic lit with IR LEDs) to diffused illumination (front or rear lit with IR LEDs). So far the results are promising. I've uploaded a video of me testing out the surface.



I think this is definately the way to go. Just looking at the video, it's clear to see I get fairly accurate results using only ambient light. With a rear lit setup, I'm sure I can make this work even better. That's it for today.

Oh, and Happy Birthday Nation... I got you a giftcard.