I was wondering if front illumination could work in all environments. For example, what if infrared LEDs are placed within and around the entire frame (slightly above) whatever clear surface you use (glass, acrylic, etc.). They’d hopefully be hidden from sight, but would shine light right above the surface. Basically it would be similar to how we place LEDs for FTIR, but instead of shining into the acrylic, you have it shine above (on top of) it. When you place something on the surface it would create dark/shadow spot I would assume.

I’m wondering if the light source needs to be far away (not so close to the surface) and whether it needs to be centered above the surface, rather then angled towards center from the sides of the surface. I’m wondering if you would get less contrast in the center than the edges or not.

If something like this worked, I would think rear-illumination would be unnecessary (it’s basically the opposite), and also there wouldn’t be a need to switch from rear-illumination in a dark environment to a front-illumination (sun or lamp) in a light environment.

Any thoughts?

MY guess is it would run the same as normal FTIR.. You’d probably be better off with either using rear illumination or an extremely powerful set of IR lights for front illumination. I’m just guessing here though, I’ve never tried front illumination..

Yeah, I guess I may have posted this a bit early lol. I should have done some tests since everything will be assumptions. Whitenoise did some tests with front illumination working with touchlib but the light source isn’t exactly close to the surface, so I was just thinking if it’d be possible to move the light source closer to the table, and then close enough to be a part of the table (shinning straight above). The LEDs would need to be powerful enough to bring light to the center of the table which I think may be possible with a table width less than 2-3 feet (.6m).

Yeah, Whitenoise’s experiment was with the acrylic up against the window.. I don’t think he used any IR LEDs in that situation, kinda like using Christmas lights instead of LEDs.. Another alternative though to having a smaller table would be to hover an IR lamp (search ebay) about 2.5’ over the top of the table, but the downside is, depending on how big it is, you might need two lamps, maybe even three, and even on ebay these lamps can cost about $30 USD a piece…

Well we’ve tried both and we’re still not sure which method to use.. However rear-illumination has more pro’s than cons. We experienced that the sun is so strong that artificial light must be real strong to make any difference. An other con for front-illumination is that shadows make detection of fingers real difficult (standing for the screen occults the light making fingers real hard to detect -> because they are in your shadow). Putting leds near the surface didn’t do the trick (real fast try) for us -> not strong enough, at least it couldn’t beat sunlight whatsoever. So i guess rear-ill. is ideal..and it looks better :D

Ok, so you did kind of try what I mentioned. LEDs near the surface.  I guess the light source would need to be strong enough to cover the entire surface equally which isn’t easy to do from the sides (instead of the center or dispersed throughout). Also I would think if the light is right above the surface (and not in back of the person using it) it would eliminate the problem of casting a body shadow.

I’m trying to think of something that will work in all lighting conditions.

"Also I would think if the light is right above the surface (and not in back of the person using it) it would eliminate the problem of casting a body shadow. “

excactly what i thought but it didn’t work out for me -at least not in a room with sunlight- but maybe with more and stronger leds? I hope you figure something out…

At least we’re thinking on the same wavelength lol. I’ll see if I can do some experiments with this. I think rear-illumination has a real advantage over ftir especially since it eliminates compliant surface problems and allows for other objects to be used.

I’m thinking if front illumination has a reasonably effective way to fit like we’re thinking, it may prove to be an even better solution. Although, I’m also thinking of using fiducial markers, and that probably won’t work with a font illumination setting since the bottom of the object would need light :-\