can bare acrylic work?  It’s must use compliant surface ,right?

You dont *need* a compliant surface but then you will have the classic problems of finger grease and friction reducing dragging operations

You prolly want to use really thing acrylic or PC, and then add a thin layer of silicone rubber on that.  The thinner the plastic, the less parallax.

as far as i understand you can use High Power LED to make enough light to replace the projector, isn’t it? I think it will possibal that we can replace the projector by High Power LEDs. Cos LCD can display all the image. Just need some high light to make it lighter. And the LED can emit light but not hot, so it’s the best choice I think.

PS:good job! nima. I’m also cant wait to setup LCD FTIR system!!

Hi Nima, great job so far!
Are you planning to make this a DI one? Think that’ll rock.
Was also hoping you could make a how-to of the LCD screen part.

Keep up the good work!

This version is a type of DI.... which I will end up using because I’m going to do object and finger tracking.

The monitor I used was 19” LG monitor that was once my computer monitor.  I’m planning on getting a LCD TV (hopefully a 37") for the next prototype.

I’m also on planning on putting together a how-to for LCD monitors too once I return back to school after the break.

-n

Nima, since you’re really doing a front illumination setup, does fiducials/objects track with this? I didn’t think front illumination would work for fiducial tracking.

yeah it’s possible!!!

; you can not use reactivision fiducials because the organic-patterns will be interperted as finger points.  But touchlib searches for rectangles so you can have square-fiducials alongside finger points.

Before I left for xmas vacation, I was able to get finger points and objects *recognition* going… hopefully some of the touchlib programmers can enable the pumping of fiducial info through OSC so we can use the data.

Here is an image of a hand with an object with a square-in-square fiducial.

When I get back into town; I’ll post info and pics on how to exactly do this (it’s a simple concept actually, instead of reflecting IR light, the object is actually emitting IR light in a square pattern.... I have three objects so far, each with three different square patterns, and there is an IR LED inside each object facing down.  The IR light shines through a mask with squares cut out)

I first tried this method with FTIR a few months ago but it wasnt as crisp because of the diffuser-projector screen. 

-n

i forgot to mention:

currently, if touchlib finds a square-pattern, it will assign it an ID #.

The console will spit out “Square Detected: 5” for example, and another fiducial will report “Square Detected: 11” to the console

OK yeah. I was thinking reactivision fiducials. Didn’t think you could recognize fiducials on paper using front DI. It shouldn’t be too tough to have someone enable sending the ID and info through OSC.

this is pretty cool nima. i think the biggest problem is with all our setups that all the time we have no real physic/ tactile / haptic feedback. do you think you could implement this kit from immerson to your system?!
http://www.immersion.com/industrial/touchscreen/integration.php

this is *very!* interesting.....  Yes; it can be mounted in LCD MT systems because the actuators are installed around the bezel of the display.  So, you dont get precise location haptics but it looks really promising actually.  Have you ever experienced this display?

After looking at the tech. sheet, this concept can be DIYed.  I need to do more research to see exactly what type of motor they are, but once that is figured out, it should be reasonable to make a microcontroller board for the motors.

They claim that they can create different haptics for shape, duration, frequency, etc… all of which are easy to implement through a microchip…

I played around (very briefly) with haptics.... not professional haptic/tactile displays, more from a DIY perspective experimenting with motors… it is somewhat complicated (for me atleast since im not trained or educated in computer engineering/programming) but a system like this looks reasonable if the proper motor is located.

p.s.

http://www.tactiva.com/

here is another company claiming a multi-touch and tactile system.... I dont have much to say about their product because a) im not convinced that they have another more than a concept because there is no image of the actual product and b) the overhead camera seems to be obstructive and not “user friendly”

unfortunately i havent the chance to get this kit, i have contacted immersion to get one. i ll let you know.
my ftir based mt is designed for deafblind peoples so i was looking for tactile feedback. i also found an intersting patent from westerman, the former ceo of fingerworks, which was bought by apple in 2005.
us patent#:20070247429 “KEYSTROKE TACTILITY ARRANGEMENT ON A SMOOTH TOUCH SURFACE”

looks like apple is coming out with some multitouch tactile keyboards ^^

I’m still wondering what’s the best way to backlight the LCD.

Please take a look at my scheme: http://www.michaelscofield.nl/FTIR_LCD_SCHEME_1.png
Qeustion: Would the mirror disturb the the light source, or will I still have a fine working backlight.
And anything else that’s not ok in my scheme?

Please let me know, thanks.

I’m pretty sure it would disturb your light source. Even if you use the original backlight system (from the screen), I think that the mirror will disturb a little (maybe using a light filter would help).
But I like your idea, a mirror is a good way to avoid big boxes and that’s what I’m looking for.