Multigesture.net

Last week I managed to get touchlib compiled under Ubuntu Linux (thanks Xvarg). While bug fixing and updating my touch tracer source code, I also tried to compile it in Ubuntu. Since I am not a linux guru, getting the code compiled took quite some time (linking libs, changing cflags). After fixing a few compiler warnings and errors (g++ seems to be more strict) I managed to get touch tracer compiled.

It is nice to see that touch tracer is actually running smoother/faster in linux than in windows. Obviously my Intel Extreme Graphics 2 (aka Intel Extreme Crap) is not a high end videocard, but where Windows does 40-50 fps on average, Linux is able to boost it up with 10 fps extra.

In the video below you will notice a few glitches. These are not caused by touch tracer (really :D)!. My FTIR panel seems to ‘leak’ light on the edges. After creating some demo movies I noticed that not only my finger tips where reflecting the light. Also other parts seems to reflect (just enough) IR light to fool touchlib. It’s even possible to create blobs when you hover your hand near the edges. A possible solution is to extend the baffles just a bit.

Touch Tracer v0.2 example video

Changelist:
- Added tracelines
- Added funky blob colours
- Made the source linux compatible
- General code cleanup

Download:
I’ve included a windows binary and the source code of touch tracer for linux users.
By using the included Makefile you should be able to get it up running in notime.

Touch Tracer v0.2 Binary
Touch Tracer v0.2 Source code

Optional downloads
msvcr80.dll and msvcp80
- Required if no vs2k5 runtime dlls are installed
Visual Studio 2005 SP1 Redist. Package
- Entire package if the one above fails

It works!

After the last blog not much had to be done about the FTIR screen. Actually I only used some tape to hide the wires on the sides.
With all LED’s in place and the frame being placed on two chairs it was time to finally test the FTIR screen.

All LED’s are fully functioning !

Testing with 5 and 2 fingers…

10 finger test:

Video containing a summary of the project and some demo applications

To get the Touch Tracer tool (from the demo video) please click on read more.
Read the rest of this entry »

In this final step of building the FTIR prototype I will cover the soldering of LEDs.
In my previous blog post I already calculated that 30 LED’s would be enough to cover the entire plate.

As power source for my LED’s I will be using a 12V DC Adapter (1000 mA).
To calculate the LED array and the amount of needed resistors I used an internet tool the LED Wizard.

According to the wizard I should create 5 array’s of LED’s. Each array should contain 6 LED’s and one resistor of 68 ohms.

Click the link below to continue reading.
Read the rest of this entry »

Inspired by other FTIR prototypes I decided to use an aluminum frame. The main reasons for are because its cheap and strong.

Before actually working on the frame, I made a drawing in AutoCAD. Within this drawing I could easily calculate the amount of IR LED’s I would need and at the same time see how the IR light beams would spread.

FTIR Prototype Drawing

The local tool shop was friendly enough to cut the aluminum frame (H shaped) into the sizes of my drawing.
So at home I just needed to remove the edges in 45 degrees.

Click the link below to continue reading.
Read the rest of this entry »

The FTIR prototype which I am building will be using an acrylic plate which will be mounted in a frame.

I tried to maintain the aspect ratio of a regular screen (4:3). The size of the plate is 41 x 31 cm and about 8 mm thick. The actual area which will be captured is 40 x 30 cm.

Two days after ordering the acrylic plate I received it in a nice package:

In the image below we can see the plate has been cut roughly.

(The plate looks a bit blue but thats due the colour of the protective layer)

Because we want to beam IR light into the sides of the acrylic plate, it is necersaily to polish the sides.

Click the link below to continue reading about how to treat the acrylic plate.
Read the rest of this entry »

For the past several days, I have received parts for constructing my touchscreen prototype which is based on Frustrated Total Internal Reflection (FTIR).

Today I will cover how the lens of the Philips webcam can be dismounted and replaced by one without an infrared (IR) blocking filter.

Most consumer webcams contain an IR filter which will prevent IR light coming through. This is done to prevent image distortion. However in our case we want to capture the IR light. On some webcams it’s possible to remove the IR filter (which is behind the lens and often has a red colour). With our current webcam from Philips this isn’t possible. Therefor a new camera lens has been ordered on eBay.

It was called “4.3MM 4.3 MM CCTV Camera board Lens IR DVR NEW”

Some specifications of the new camera lens:
* Glass Lens
* Focal length: 4.3 mm
* Aperture F/1.8
* Image format 1/3 inch
* Angel of view 68°
* Mount M12 x P0.5
* Size ?14×14.75

Click the link below to continue reading about how to replace the camera lens.
Read the rest of this entry »

Perhaps you noticed an article about multi-touch screen devices which I posted sometime ago.

Inspired by the movies and publications of Jeff Han and Mitsubishi (Diamond touch table) I will be doing my master thesis research project on multi-touch gesture recognition and processing.

At this moment I do not have a multi-touch panel yet, hopefully at the end of this month I will have a working prototype.

A Philips SPC900NC webcam which will be modified for multi-touch project.