For the Project I was tasked to continue reasearch and development on the EVL's Omega table. The Omega table is a multi touch project where the user will be able to interact with a smooth surface using only finger touches. The screen will also be in stero giving a immersive 3d view into the generated world.
Progress
When I started this process the table hardware was built. There was a
frame, the acrylic was already cut and in place with LEDs, and the camera
was already mounted. The major issue that I was tasked with was
increasing the signal strength, because we have to transmit the IR
reflections through multiple layers. I descovered that the camera that we
were using, Natural Point Optitrack SLIM v100, had a exposer setting that
made our finger reflections visable through the natural Point grey scale
demo. Through tweaking the configuration settings I was able to
demonstrate that we are in fact able to produce a strong enough signal
through two layers of LCD and a defusers.
Now that I knew that our design was able to reach our goal I decided that
we needed demo software to run on the Omega table for further testing.
After some research I descovered a common touch liberary called Touchlib.
Touchlib is a open
source project/liberary that has a very active development community.
Initially I was unable to get the optitrack camera working. I discovered
that the Natural Point camera's use specific driver signing that is not
compatable with Touchlib. Natural Point offers a direct show plugin that
allows the camrea to be seen as a standerard webcam, but I was unable to
get that feature working fully. I tested a logitech web cam with the
Touchlib liberary and it worked. I was able to inreract with the smoke
demo program using one LCD that was turned off. I discovered that the
moniter that we are using only supports a digital i/o so I requested a
computer that supports this.
One of the ways we tryed to increase the IR reflection strength was to
apply a think layer of silicon. At first it seemed to be a success in
increasing the signal strength over all. I was a 5-10% increase in
strength with my hands. However, we have a someone in the lab with
very dry hands. His hands were invisable before and now they show as
strong as mine. After we applyed the silicon the surface became very
sticky. I belive this was due to the plastic covering I was using to keep
it clean. I removed the covering for about a week and it became less
sticky, but more dusty. It finaly got to a point were I could almost
drag my fingers accross normaly. I was testing draging with the smoke
demo and the silicon tore in the center. I belive that for our project
silicon is not a pratical way of increasing signal strength.Images from
silicon process.
Progress 6/3/08
I curently am waiting on a computer request to be filled. I also am in negoation with a acrylic vender. I also am preparing for a demo on the 27 for the Science Museum of Minnesota, where we'd like to have the rain table softwere running on the Omega Table.
Progress 6/5/08
I have been doing allot more research on how people have been able to achieve the FTIR effect with a very strong signal. After taking a step back I decided to pursue a simple idea. Would polishing the edges improve the signal strength. First I practiced the polishing process on a scrap piece of acrylic. This turned very successful. I then began on one of the 45* angle sides. I was able to improve the visible clarity, giving it a glossy finish. I also tested removing scratches from the surface. I determined that I was able to both polish to a glossy finish and remove scratches. But this process is very time consuming. I tried to gauge the results through testing. I was unable to view a noticeable difference. I also received the computer I requested so now I will be able to do testing more efficiently.
Progress 6/6/08
My goal for today is to get the new workstation working with the camera, so I can display it on the LCD screen. I remembered, to properly view the LCD screen the diffusers needed to be in place. I put the thick diffuser on the backlight panel. This caused the screen to be visible but you had to be looking directly down into the light, on the sides you can see the walls. I remembered that to fix this problem we have a thin diffuser that we place right against the LCD. I tested this and it works, but I needed a way to hold it up. I came up with a idea to build a frame for the thin diffuser, building it with some extra foam core strips. I taped the sides of the diffuser about 1 cm on all sides and had a strong under frame. This held the sides of the diffuser against the LCD while the center sagged a little. But the center was already good so this had the effect I was looking for. I transferred the camera/tracking software from my laptop to the new workstation. I was able to the optitrack grayscale demo working but the Touchlib software is still missing some dependencies that I need to get online. I am going to wait until I get internet access to do this. I modified the camera mount to allow for vertical movement because I was testing how the exposer of light to the camera lenses effected the image. At the end of the day we received the shipment of the Strong LED's we ordered. They are about three times the size of the original ones. After consulting a friend that works with electrical circuits I decided to create our circuit on Monday with his help.
Progress 6/9/08
Today I did more software updates to the workstation. I was able to get internet access and continued to get all
the libraries working again. We also received and installed the physics card into the workstation.
I downloaded some physics demos and found a fabric demo that would fit perfectly for
the software we are planning on developing. I also worked on building the
circuit for
the larger LED's we ordered. We created a powerful circuit but we did not
have enough resisters to complete all four sides. We ran some tests with
our strong LED prototype and compared it with the original LED's. We found
that the larger LEDs are stronger.
also we determined that we currently are not using the most efficient angle to light the
acrylic. On a side note, I discovered that if instead of using the FTIR effect, if we
raise the LED's just above the surface we get a incredibly strong signal even with the
backlight and LCD on. This may be a option if we cannot generate a strong enough
signal through FTIR.
Progress 6/11/08
Today I drilled the holes into the short side plastic frame. I also completed the circuit for the prototype large LED array. I am testing using foil tape to reflect the unused IR light back into the acrylic to see if this makes any differnece.
Progress 6/12/08
Since the resistors were on the way I decided to prepare the acrylic for 4 sides of LED's this meant that I needed to polish the two short sides. I also preformed more tests with the cameras and touchlib. We are approaching the point where I need to start to program so I looked into the physX API. I also discovered that the resolution of the screen seems to be software dependent. Which re-enforces my belief that I need to start making some home bru apps.
Progress 6/13/08
After more research I discovered that the Dark Basic physics engine, Dark Physics natively supports the physX API. I installed dark basic and wrote a few samples to get familiar. I believe that with Dark Basic I can meet the demands of the software I am planning.
Progress 6/16/08
The resistors have arrived. They are about half the size but match our specc's. To test the resistors I completed the circuit on our prototype. They did not work. After a TON of testing and troubleshooting I discovered that it was a bad LED. And that the proper voltage setting on the voltmeter is the second click. I replaced the bad LED and it worked. I finished both long side LED arrays and they both are working good.
Progress 6/17/08
At today’s tech meeting Julio Obelleiro gave a presentation of his game engine that he is developing. This engine supports standard game engine functionality and camera tracking. I was very impressed with this and decided to look into it. After playing with it I decided that with the camera tracking support Julio's Vision engine would be a better choice. I also was able to capture this stereotypical image of touch screen tracking.
Progress 6/18/08
Today I build the power leads for the two leads for the two short side LED arrays. I used the power connecters off a computers power supply. I also used the red and black wires from the power supply. I soldering the new power leads to the existing ones, creating a similar circuit. I decided to change the build of the short LED arrays slightly. We only have 20 large LEDs left. So im using 5 per side for the short sides and then two series of 6 small LEDs. This uses less large LEDs and fits the frame better that only has 17 holes.
Progress 6/20/08
We finally received a new piece of acrylic. It is the correct size, with all four sides cut at 45* angle, and polished. ATM I am placing it on rubber supports but id like to find a more permanent solution. I came up with a frame design for this where the frame would slide over the entire frame, and be secured with some type of bolt system. I need to talk to support to determine if this is possible some how. I ran a few tests and I believe that the results are brighter than before. Also I was able to test the reflective foil. It DOES help in reflecting the IR. I talked to Jason and we agreed that in future implementations we will get some reflective spray paint to do this. Iv been talking with Khairi to get the rain table software working on the Omega Table. He is currently trying to reconfigure the workstation to get the tracking software working.
Progress 6/27/08
IT WORKS!!
After long hours in the office all week. A ton of help with the software from Khairi. The Omega Table is running the
Rain Table software. We had a meeting yesterday that disused the future of the table. We are going to finish the stereo
on the 30in screen. Build a couple more proto types and demo able versions. Then go bigger and better. We also had a meeting
with Minnesota. They loved the table and want us to build them one.
Current objectives:
switch to optitrack tracking API
re-build LED array with lens and fixed power leads
multi camera
A tech side note: the small LEDs circuit is 8 in series with one 1/4 resistor. Also I am going to prototype a new LED array design
for this version im going to take a cylinder tube(a hose or poster tube), spray with chrome paint. and cut so it fits tight around the
acrylic. hopefully this will produce a extremely bright signal without too much bleeding.
