detect where my hands touches my glass tables surface

Go To Last Post
22 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hey guys :D

I have a glass table which have a white layer on the bottomside (paint of some sort) LEDs can shine through it.

I'm thinking of making it in to some sort of control surface... so I ex. can tap at one location and then it means start etc.

What is the cheapest way to do this ?
the resolution doesnt need to be more precise than 10x10 cm (surface is 70x120cm)
I have an AtMega32 right now for this project :)

//B4Me

uC's: Atmega16, 32, 64, 128 and Attiny13
Lang.: C
Interests: Small scale robots AND sensor monitoring system

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you only need a few tap points how about using a few phototransistors stuck onto the underside of the glass. All you have to do is detect when they are darker than a set threshold below the brightest of them. That means you hand is over that tranny and blocking the light.

If you wanted to map the location of the tap you could do it with three microphones. ADC and have a SW routine that detects a taps. The relative volume between the channels and a bit of trig gives you the XY location.

You could also do it with, electrostatic fields, RF, magnets, smoke AND mirrors. As you said tap I'd prob give the audio solution a go.

I hope this is a start, I'll be interested to see what works :)

<º))))><

I am only one lab accident away from becoming a super villain.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Capacitive ... proximity ... QTouch. Think theremin.

Cheers,

Ross

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I dont think time-of-flight of the 'tap' to the mics works unless you have simultaneous sampling?

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

Capacitive ... proximity ... QTouch.

Yep, when you go through the Qtouch info you will find examples of making large-scale "panels", both with discrete "buttons" and "slider" types. I don't remember an X-Y offhand.

The pad design is a bit of a black art such as RF antennas, but bigger does tend to be easier in our limited work.

DDR? http://en.wikipedia.org/wiki/Dan...

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I was thinking of just relative max volume of the detected taps (maybe a bit of calibration would also needed to sort out the position on the table).

It might be possible to get time-of-flight working, if the sample rate was high enough. Possibly use 4 mikes and do 4 XY calculations using the 4 diff sets of 3 variables. You could then see how well they agreed with each other.

Instead of ADCing the mikes, maybe just use them as 1 or 2 bit (via resistors) and tag them onto an 8 bit port to give you the simultaneous sampling. I am thinking of those tiny little FET mikes that include an opamp to give gain and buffer the output, but it may be better to have a separate circuit that could also include a BPF and a peak detector. Just thinking out loud etc

<º))))><

I am only one lab accident away from becoming a super villain.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thats like a logic analyzer... just start sampling 4 inputs on portx and note the count that each input goes hi. Clever. I was imagining one of those blacklights like on CSI and look for greasy fingerprints on the glass..... haha

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks for all the good ideas guys :D
like the last CSI ref ;)

uC's: Atmega16, 32, 64, 128 and Attiny13
Lang.: C
Interests: Small scale robots AND sensor monitoring system

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What about an IR matrix, like what is sometimes done to make a touch-enabled screen out of a flatscreen.

http://mortalspaces.com/diytouch... (see BEAM ARRAY)
http://www.tomshardware.com/news...
http://www.maximumpc.com/article...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Once, I saw a 'touch panel' done with a led matrix array, where the very same LED's were used as touch sensors, but I can't remember where I found it, and I don't have the time to search for it again, but it seems to me that this could be used for your purposes.

Guillem.
"Common sense is the least common of the senses" Anonymous.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

valusoft wrote:
Capacitive ... proximity ... QTouch. Think theremin.

Cheers,

Ross


I'm with Ross on this one.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Do you have "things" at the table. If not place a pressure sensor at each corner, perhaps it could work anyway with offset.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

It's my normal couch table, so yes there will from time to time be placed various items on the table... the software should then ignore them if they stay the same for X time.

Underneath the table I have a LED matrix, which have 16x32 pixels... I can control these ON/Off individually, if this could help ?
The LEDs have there own uC / "controller" and another uC collects data (temp., time / date, LDR, Serial bus) and sends commandoes to the LED_controller. Both are AtMega32.
The last uC is the one I would use for detection also.

uC's: Atmega16, 32, 64, 128 and Attiny13
Lang.: C
Interests: Small scale robots AND sensor monitoring system

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

google FTIR multitouch.

Basically a web-cam and some IR LED's and you're good to go.

Writing code is like having sex.... make one little mistake, and you're supporting it for life.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry... how would you handle the image-processing from the webcam with one AtMega32 ?

uC's: Atmega16, 32, 64, 128 and Attiny13
Lang.: C
Interests: Small scale robots AND sensor monitoring system

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

google AVRcam

Also check out the EE4760 projects over at Cornell.

Writing code is like having sex.... make one little mistake, and you're supporting it for life.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

But then again.... I only have 4 cm of free space between the glass surface my LED matrix :-/
so I cant use this idea :(

uC's: Atmega16, 32, 64, 128 and Attiny13
Lang.: C
Interests: Small scale robots AND sensor monitoring system

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you build your own matrix , this should be possible
http://cs.nyu.edu/~jhan/ledtouch...

http://projects.dimension-x.net/...

http://www.touchuserinterface.co...

But i'd try to test some Capacitive-sensor first , to see if it detects through glass.
http://elm-chan.org/works/capsen... or Atmels touch-lib.

But the led stuff would be "nicer to look at" , as you can prob. see the construction through the glass.
Maybe you can make a grid of thin wires as capsensors

/Bingo

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

There shold be no reason why a capacitive through glass sensor ought not work.

Back in '77 we were looking at just that kind of sensor for some stove top controls.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ignoramus wrote:

Back in '77 we were looking at just that kind of sensor for some stove top controls.

But that's "Cheating" ... You could do detection with a mic ...
In 77' the person would say "Outchh" touching the hot "radio-tube" :-)

/Bingo

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ahh..but the thing is..in air "ouch" travels at 300m/s whereas in glass it is close to 5600m/s.
So a nod is as good as a winkto a deaf microphone.

I as refering to capacitive method based on thumb print sized adjacent electrodes attached to rear of glass panel.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The Nov 2008 issue of Circuit Cellar Magazine has an article on the time of flight concept using 4 mics in a plastic table top.