Ask Hack a Day: Touch Screen Hack

Reader [Chad Essley] asked us:

“I’m wondering if the vast knowledge base of HackADay’ers out there might know of some way to turn almost any laptop into a touch screen of some kind. Actually, any surface.”

He has an older Wacom Tablet, and would like to be able to add resistive touch screen capabilities so that he isn’t forced to use the Wacom pen. Being an artist and part time hacker, he even summed up the question in a comic-style post.

We know of tools such as the EZscreen instant touch:

which allow “Clip on Tablet abilities”, however you have to use their pen, which doesnt help with the tablet problem, but may help with your “any monitor” problem. Another solution is to buy a generic touch screen kit (usually on eBay from somewhere in China), and install it yourself. That will work for most laptops and desktop monitors, however we wouldnt recommend using it with your Wacom tablet, as it would probably interfere with the pen functionality. Hak5 recently did an episode on Touchscreen kits, but there are also a number of places all over the internet to find step-by-step how-tos for kits.
We dont think you could use an infrared camera from the side like in your drawing, because it would be very difficult to calibrate (wouldnt be able to tell the difference between cold hands, and hands that are farther away/etc in our opinions). Things like the Microsoft Surface use infrared, but from underneath (which is much easier for machine vision to handle.) A flat, unlit surface (not like a monitor or a tablet) could use something similar to the laser keyboard.
We have also covered a couple of gesture and touch based input tools which may help out anyone willing to take on this task. TISCH is a top-down camera based multi-touch framework, and Scratch Input is a tool for using acoustic signals on a wall or a table to interpret touch and drag motions.
Unfortunately, that is about all the ideas we could come up with, but the Wikipedia page on Wacom tablets might be a good place to start for background information on what could or could not work with this configuration. Hopefully some of our ideas will spark the interests and specialties of our readers, feel free to respond in the comments, or send any ideas to our tip line!
Thanks again to [Chad] for his question, and feel free to send any other ideas for Ask Hack a Day.

Comments

  1. BlackOp333 says:

    Use a wiimote

  2. mickern says:

    What about some sort of radar/sonar system? Would it work on such small distances?

  3. Miklos says:

    Something like the EPOS digital pen? Some of these digital pens can also be used as mice..

  4. floe says:

    Pretty similar to our research paper from ITS 2009 this year, see video: http://www.youtube.com/watch?v=ARvtruxJDH4

  5. medix says:

    @mickern: Probably not. Ultrasonics really aren’t useful at close range. You end up ‘hearing’ the same ultrasonic pulse that was just ‘sent’.

    His camera idea is probably a good start, but you’d most likely need two of them for triangulation. Not sure really..

  6. dbear says:

    Two cameras with either a line aperture created with a slotted lens or the same simulated in software. Although I’m not a software guy it seems the software should be pretty simple.

  7. strages says:

    It can be done with an IR source and an edge mounted webcam. I did something similar to my laptop once using Touchlib and an 780nm laser with a 120 degree line lense.

    Seeing this post actually makes me want to revisit the idea and try and streamline it. Thanks!

  8. strages says:

    Also, you may want to use more than one camera at different angles, if doing multitouch because of occlusion by other fingers in the field of view.

  9. Rob says:

    The SISO Tablo still requires a pen, but can be added to any laptop:

    http://www.hantech.ca/product/product_detail.asp?M&S=3&bCateName=TABLO&CATEID=00030001

    You could also buy a resistive panel, but the additional thickness might interfere with your digitizer’s calibration

  10. cgmark says:

    For the wacom tablet you would be better to tape the pen to your finger or make something from the electronics inside the pen to fit on a finger. Wacom uses RF to both power the pen and track its movement.

    I would scrap the wacom tablet as a base for a touchscreen and start with creating one from scratch.

    If anyone got in on the ST discovery kits for $7 the other day, they come with a touch interface and the libraries for expanding it to larger areas. Really the best way to do this is the way it is currently done with a touch panel .

  11. nnx says:

    I was thinking on making one similar to the wacoms.
    pen has 3 coils, 1 feeder coil, resonantly coupled to one coil at the tabled, which sends energy to the 2 other coils.
    the two coils (or mini-antennas) simply emmit rf-radiation at a specific frequency.
    at the tablet there are 6 receiver coils, 3 each tuned to 1 in the pen, they triangulate the position of the pen, since there are 2 it is possible to describe not only the position of the pen but also its inclination, with a little script it could be possible to select a appropriate brush (round-eliptic) in photoshop or similar programs.
    It would also work in 3D, making it ideal for modelling in 3ds max or similar program. to change the intensity of the brush the (metalic) tip of the pen could be pressed into one of the coils, changing its inductivity and thus its frequency.
    Of course 3D movement detection requires one order of magnitude more data than 2D, so maybe one could apply a crude on/off resistive foil to the screen so that the position of the coils is only scanned when the pen touches the surface.
    (sorry for my bad english, im from germany. Sorry for the incoherent style, it was just a random idea in my head)

  12. The Steven says:

    Step One:
    Assume it can’t be done.

    Step Two:
    Tell all your contacts it can’t be done.

    Step Three:
    Sit back and wait for someone to prove you wrong.

    Step Four:
    Offer the person who makes it workable some money to build one for you.

  13. Michael L. says:

    I know this might sound like a dumb idea to most of you but you could gut the wacom tablet and build a transparent enclosure for it, then use the IR camera idea (undermounted/bottom mounted, however you want to say it) in the center of the enclosure. I personally don’t know how the wacom tablet’s guts are configured so this may or may not be a viable idea. If the receiver circuitry is just in the top then this should work fine. If it’s just a few wires running through the tablet you might have some “black” areas, areas where the signal doesn’t show because of the blockage from wires, but they should be minimal. If the guts are all housed in the “drawing” portion of the tablet it probably would be easier to go with a touchscreen kit, as others have said before.

  14. James Munns says:

    @The Steven:
    Step 5: ???

    Step 6: PROFIT!!!

  15. maxster says:

    So, is this the business man ask a hacker?

  16. Chad Essley says:

    The webcam / laser approach is closest to my idea. I’m not after a way to create a brand new screen, rather to make any existing surface touch capable. My desk for instance.. the cat! (not really) -but something flat. So an enclosure is right out. @strages – Your idea seems the closest. Anyway, thanks to a very cool community on HackADay. I’m sure someone out there will build this thing and 6: PROFIT! I’d be right in line to buy one.. :-)
    But hey, I’m only a cartoonist..

    Best
    C

  17. strages says:

    @Chad thanks, but I can’t take complete credit. This is how a prototype Microsoft multitouch wall works. 3 IR lasers at different angles placed at the bottom of a large pane of glass/acrylic with line lenses. A Camera sensitive to IR is places at the top and software handles the blog detection and manipulation. In lue of the camera, as @BlackOp333 said, you can use a wiimote and it’ll do blob detection for up to four points.

  18. mutant says:

    At my school, the special class has something attached to the monitor that makes it become a touch screen, that way its easier for the kids to use. It was attached via velcro to the top and had two thin panels that extended across the top and right hand side. From the looks of it.. it’s some sort of theremin used as an input device. I don’t recall the brand name, but i got to mess with it.. it was pretty cool.

  19. sam says:

    I would just like to point out that the Microsoft Surface and nearly any other IR based touch system uses near infrared, like in your remote control, not far infrared which is heat and the like.

  20. Rob says:

    Engadget just put a short article on Motorola investing an an acoustic multitouch technology:

    http://www.engadget.com/2009/12/03/motorola-invests-in-anywhere-multitouch-technology/

  21. vicx says:

    The hacking for this is already done. Of course you can design your own solution but this is a good place to review the progress already made in diy multi-touch and related technology. http://nuigroup.com/forums

  22. olh says:

    You just need two perpendicular axis of an electromagnetic field inductor, maped to your table/screen. Just like a themerin.

  23. olh says:

    theremin*

  24. ltkerr0r says:

    Hak5 did a brief segment on adding a touchscreen to a LCD.

  25. Steve Z says:

    that Hak5 video was excruciating to watch.

  26. blizzarddemon says:

    Johnny Lee’s wiimote hacks to the rescue! http://johnnylee.net/projects/wii/

  27. misterdumguy says:

    so the two styles of touchscreen are capacitive and resistive? anyway, if conductivity is the requirement, how about building a solution that consists of a thin, conductive, transparent film (idk, that really thin gold foil, something like that, anything thin that lets light through) a plastic overlay, like maybe what teachers draw on to use with a projector.(but perhaps something thinner, softer) put those two together, and figure out how to make them not touch until you touch them. one complicated solution for the last requirement is to fill the space between with a thin layer of a nonconductive liquid/gel so as not to interfere with optical transmission, but when you press on it, the conductive film touches the touchscreen in that area only.
    yea thats really friggin complicated eh?>

  28. dan says:

    wacom screens place the wacom tablet below the LCDs, the range on a wacom is very good, up to a cm to two. They can be calibrated at any height.
    Adding a touch-resistive panel over a wacom will work fine, and some tablet PCs do implement both, so you can tap with fingers, or write with a wacom pen if you want more accuracy.

    Turning any surface into a tablet is another problem. If you only want tap sensitivity, the pizzo-pickup based tap thing works fine to triangulate to ~ 1cm accuracy, just using a decent soundcard. If you’re using a large surface, that kinda accuracy may be ok.
    That won’t work for dragging though, you’ll need a camera-based thing like others have suggested.

  29. Miklos says:

    “Ultrasonics really aren’t useful at close range. You end up ‘hearing’ the same ultrasonic pulse that was just ’sent’.”
    Rough calculation. 1 cycle of 1 MHz pulse = 1us. Speed of sound is 0.33 mm/us. That corresponds to 0.33 mm propagation distance. So interference between the transmitted pulse and scattered pulse from pen tip is only expected to be a problem for objects within this distance. Even at lower frequencies, this distance is quite minimal (20kHz->1.65 cm..I’m sure you can place the device accordingly). However, this distance will also give the resolution of localization, so short, high-frequency pulses are preferable.

  30. Miklos says:

    Correction to last post:
    interference between the transmitted pulse and scattered pulse from pen tip is only expected to be a problem if the distance between the transceiver and pen tip is < 0.33 mm

  31. sina says:

    hey guyzz …. i really liked the discussion and the whole idea … but i was woundering if any one could help me in finding a way to have an input method that uses laser or lights or something not solid to input ….
    example : having an access to a screen that doesn’t has any base (like we know the flat solid surface behind it )

    any idea is welcomed …. thanks a lot

  32. jsngrimm says:

    i know this is an old post but if u dont mind using a laser pointer then u can track the dot with a clip on webcam :) theres a post on hacknmod i think

  33. dan says:

    re 1 MHz ultrasound pulse / 0.33 mm.

    you’ll have to send a pulse longer than one cycle.

    1950’s radar. pulse long enough to have some robustness, short enough to get out ot the way of quick echo from close objects. PRR, Pulse Repetition Rate, slow enough to give long range, so echo gets back before next pulse. [ limit selectable 30 - 200 miles, airborne ]

    We used X-band – around 10,000 MegaHz
    S

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 96,382 other followers