1 of 2
1
LLP Fiducials - Proof of Concept (*PIC HEAVY*)
Posted: 29 July 2008 10:59 PM   [ Ignore ]
Rank
Joined  2008-07-21
Total Posts:  21
New Member

LLP and FTIR both give great results, but unfortunately lack DI’s ability to do fiducials, a feature I find quite interesting. So, while I’m sitting here without funds to build an actual table I thought I’d test a concept I had.

Here’s my make-shift “table” which I used for this test.
The test "table"

My first idea was to try and create a sort of FTIR effect inside an acrylic “puck”, but not having any acrylic I looked around for what I did have, lots of old useless CD’s wink
After some playing around I came up with this method which produced some promising results.

1) Cut out a piece of the CD (I knew those old AOL trial CDs would be useful for something smile )
Cut a piece of CD

2) Scratch up the surface of the CD (The data side, 60 grit sand paper with light pressure circular motion works quite well)
Scratch piece of CD
You should have lots of little surface scratches
Scratched piece of CD

3) Draw your symbol (Hand drawn proof of concept)
Drawing the Fiducial
The fiducial is ready (I made a couple more)
Fiducials

Placing the fiducials on the “table” you can see the edges glow from the red laser
Fiducials on table(Ooohh… pretty smile)

And the results
Results
It’s not shown here, but a bit of the laser light does make it through the CD piece to illuminate fingers behind it as well.

There’s definitely room for improvement, but for a proof of concept with my rather low-end make-shift setup I think the results are definitely promising.

Hope someone else finds this interesting, and I can’t wait to seem someone try this with more powerful setup (AlexP? wink)

Image Attachments
LLP_Fiducial_result.jpgLLP_Fiducial_6.jpgLLP_Fiducial_5.jpgLLP_Fiducial_4.jpgLLP_Fiducial_3.jpgLLP_Fiducial_2.jpgLLP_Fiducial_1.jpg
Profile
 
 
Posted: 29 July 2008 11:32 PM   [ Ignore ]   [ # 1 ]
Avatar
RankRankRankRank
Joined  2008-02-12
Total Posts:  1120
Member

Very Interesting… Nice idea… i dont know how it will work with FTIR though i can see it working with LLP since LLP has a IR light plane.....

 Signature 

Company Website - Knode Research and Development
My Youtube

Profile
 
 
Posted: 29 July 2008 11:42 PM   [ Ignore ]   [ # 2 ]
Rank
Joined  2008-07-21
Total Posts:  21
New Member

Yeah, this was intended to work with LLP, and I’m hopeing a proper table will give even better results.

I’m still trying to think of an idea for FTIR. Perhapse if fiducial was carved into something, and enough weight applied, but I’m not sure if it would give a clean image, and I don’t really have a way to test that so…

Profile
 
 
Posted: 29 July 2008 11:53 PM   [ Ignore ]   [ # 3 ]
Avatar
RankRankRankRankRankRank
Joined  2007-04-08
Total Posts:  2539
Dedicated

Very cool. I agree with taha, I don’t see how this would work with FTIR. The easiest would just be to light up a cubed fiducial with IR LED on the inside and put a mask on the bottom of the light box for the symbol. 

The llp results look pretty good. Very clever! You would def need more contrast though to make it work well and easy to track especially if the fiducial was as complex as the ones from reactivision.

 Signature 

MTmini, MTbiggie, & Audiotouch creator & Community Core Vision Co-founder

Follow on:
My Blog | Facebook | Twitter | Youtube

Profile
 
 
Posted: 30 July 2008 01:49 AM   [ Ignore ]   [ # 4 ]
Avatar
RankRankRank
Joined  2008-05-08
Total Posts:  342
Sr. Member

Great job!!!

Having fiducials is something that many people asked about when talking about LLP method.
With more lasers I think the fiducials would be brighter and easier to detect (you would have laser light entering the cd plastic from various angles, thus lighting up all of the fiducial edges).
This is exactly what I had in mind when I addressed fiducials and LLP in my post. I’m glad that someone tried. Thanks spbogie!
When I get a chance, I’ll try something similar on my setup and post the results.

~Alex

 Signature 

Computing is not about computers any more.  It is about living!

Code Laboratories
CL Studio Live - Video like never before
CL Eye Platform SDK

Profile
 
 
Posted: 30 July 2008 04:10 AM   [ Ignore ]   [ # 5 ]
Avatar
RankRankRankRank
Joined  2008-06-04
Total Posts:  565
Elite

I disscused this in another post for FTIR, but thought that it would be possible to cast the fiducial in silicon so that when it contacts the acrylic it creats blobs in the shape of the fiducial justs like potatoe prints done at primary school!  The only problem is that silcon is mighty grippy when it touches acrylic so the rotate function that is so neat about fiducials is somewhat hampered, There may however be a more slipery material that produces an FTIR effect? One for the Tinker perhaps?!

For the LLP method i love this Heath Robinson approach in using CD’s. But rightly so, a nice highly polished cube of acrylic with the fiducial printed in etched silver tape ( sticky back tin foil! ) with silver tape on the top to trap and reflect more light down would hopefully work great.

Some experimentation is required perhaps light feeding acrylic with frosted sides would work better?

 Signature 

}---)))@

V JAY!  http://www.v-lux.net LASER tut http://nuigroup.com/forums/viewthread/6407/

Architectural Technician and Building Surveys

http://www.jpgainsfordassociates.com

Profile
 
 
Posted: 30 July 2008 08:34 AM   [ Ignore ]   [ # 6 ]
Avatar
Rank
Joined  2008-05-06
Total Posts:  58
New Member

Fantastic! I was just pondering this problem the other day. Thanks for sharing your results!

Profile
 
 
Posted: 30 July 2008 11:12 AM   [ Ignore ]   [ # 7 ]
Avatar
RankRankRankRank
Joined  2008-02-22
Total Posts:  556
Elite

ajlovegrove
I kind of had the same idea w/ a hot glue gun. Making a pattern on the cardboard or cd case and using that as a symbol.

 Signature 

http://www.justinriggio.com cool mad
http://www.niceminds.com My blog

Profile
 
 
Posted: 07 August 2008 05:53 PM   [ Ignore ]   [ # 8 ]
Rank
Joined  2008-07-21
Total Posts:  21
New Member

Thanks for the comments guys. I’ve been away on vacation, but it’s nice to come back and see some interest.

Other materials are definitally worth looking into for better results. The CDs were simply what I had readily available, and functioned well as a proof of concept. They are also thin enough to possibly be used to track an object, and the shiny data layer will reflect any light that would come up through the top back down. Two things to consider when looking at alternatives.

For the contrast part, again I’m sure that there are better materials for the purpose, or better ways of getting the light out of the material. More/stronger lasers would likely help aswell. The results may even be better just by switching to the IR spectrum with a filter.

I also noticed that the pieces seemed to glow brighter and more eavenly when the light hit a corner verses a straight edge.

I would love to see this taken to the next level and tested on a proper table, but I simply don’t have the resources to do so myself.

Profile
 
 
Posted: 07 August 2008 11:15 PM   [ Ignore ]   [ # 9 ]
RankRank
Joined  2008-04-16
Total Posts:  183
Member

I don’t think it will be very simple to get the laser light from an LLP table to illuminate a fiducial, so here’s my idea:

Cast a very simple circuit into a cube of acrylic, or even just put said circuit in a pre-made box (like a 3"x3"x3" acrylic cube). Circuit would consist of a switch, a coin battery, maybe a 555 timer and an IR LED pointed downwards. At the face that the LED is facing, put your fiducial, either as a drawn-on black pattern or something else.

The table will see this face illuminated by IR light, and the pattern should be very, very clear. This would work with ANY table illumination method, and should not be hard at all to do.

 Signature 

Visit my research blog at:
http://cs.unk.edu/~webb/

Profile
 
 
Posted: 08 August 2008 05:36 AM   [ Ignore ]   [ # 10 ]
Avatar
RankRankRankRank
Joined  2008-06-04
Total Posts:  565
Elite

H4T You’re right that is by far the most versatile way to do it, self iluminating fiducials or SIF’s man we love those abrv’s.  If you had one of those label cutters for making sticky back lettering one could quite easily cut the reactivision fiducials into a sheet of black stickyback PVC. Then stick it on to the side of the cube and there are six faces!

 Signature 

}---)))@

V JAY!  http://www.v-lux.net LASER tut http://nuigroup.com/forums/viewthread/6407/

Architectural Technician and Building Surveys

http://www.jpgainsfordassociates.com

Profile
 
 
Posted: 08 August 2008 06:25 AM   [ Ignore ]   [ # 11 ]
Avatar
RankRankRank
Joined  2007-09-22
Total Posts:  263
Sr. Member

Or you could just invert the fiducial , make a stencil from it and paint it over a piece of acrylic , then have a small led above it iluminate it.

Profile
 
 
Posted: 09 August 2008 07:17 PM   [ Ignore ]   [ # 12 ]
Rank
Joined  2008-07-21
Total Posts:  21
New Member

I’m toying around with the active fiducial idea right now. I was wandering around the hardware store looking at random things(I could do so for hours wink) and came across these soft-glow LED night lights. At $6US they seemed like a great starting point. Initial results are looking promising, but I’m out of ink/toner so I’m not getting very good prints => bad contrast. I’ll get a new cartridge and post results tomorrow sometime probably.

Sorry for the delay, had a few unplanned tasks to take care of. So, without further ado…

Active fiducial from a LED nightlight (Visible spectrum proof-of-concept)

First, the original inspiration for the test.
Inspiration
This simple, cheap(2 for $6), night light gives of a nice soft, mostly even, glow from a single LED.

Simple test
WallpoweredWallpowered - On TableWallpowered - config.bat
Starting with a simple test powering off 120VAC from the wall. Looks promissing, but not very practical with that big ugly cord running off it wink.
The fiducial was carefully cut out and placed between 2 pieces of clear sticker paper. A simple transparency would probably work just as well, but I didn’t have any wink

So, it’s time to take it apart. Disassembly was a little more difficult than I originally expected because the plastic had been fused together instead of the more common simple snaps, but a good sharp knife fixes that(cut along seal and pry apart).

Profile
 
 
Posted: 13 August 2008 02:39 PM   [ Ignore ]   [ # 13 ]
Rank
Joined  2008-07-21
Total Posts:  21
New Member

Inside, there was a simple driver circuit, and this little piece of acrylic.
Acrylic - FrontAcrylic - BackAcrylic - Texture
Dimensions: 2 1/4"x2 1/4"x1/8" (5.7cm x 5.7cm x 3.2mm)
The front of the acrylic is covered with a defuser that looks/feels like a heavy tracing paper, the back has a dotted pattern to break up the internal reflections. There is a small semi-circle cut out for the LED to help spread the light through the acrylic.

LED circuit
LED circuit
I re-used the 3mm blue LED from the light for my tests. The simple circut consist of a voltage source (5VDC from computer PSU), a 5kOhm potentiometer (variable resistor), and the LED.
LED positionTesting LED
Testing the LED witht he defuser, lots of light is leaking out the sides, but it glows nicely. smile

File Attachments
photos.zip  (File Size: 1007KB - Downloads: 347)
Profile
 
 
Posted: 13 August 2008 03:45 PM   [ Ignore ]   [ # 14 ]
RankRank
Joined  2008-04-16
Total Posts:  183
Member

Wrapped in foilTesting
A little tinfoil wrap, and the symbol from before, and we have a pretty good looking fiducial there.

Testing it out on the table gives some nice results.
Config.batreacTIVision

In testing I found that the LED at rated voltage was almost too bright, and reducing it helped eliminate hot-spots. The position of the LED affects this also. If it is too close it makes a bright hot-spot right infront of it.

If using this with LLP you will want to cover the aluminum foil with some other non-reflective surface, results from illumination acrylic with laser were not much different from CDs (possibly worse). This will also create a rather large shadow on the other side, so may not be ideal for LLP setups.

These can be easily powered with a simple button cell. DealExtream has some great prices on quantity buttons cells.

 Signature 

Visit my research blog at:
http://cs.unk.edu/~webb/

Profile
 
 
Posted: 20 August 2008 01:08 AM   [ Ignore ]   [ # 15 ]
Avatar
RankRankRankRank
Joined  2008-01-04
Total Posts:  676
Moderator

This idea is WAY Cool! Nice concept spbogie! grin

 Signature 

Get Sparked at sparkon.net
ORION Multi-Touch Display (blog)
My YouTube

Profile
 
 
   
1 of 2
1