Re: FALCON Open-Source Object Recognition System Now Available!
 

For the rollover effects. If there is a non-javascript way to do that, I would be interested in knowing about it. :)

Did your mentor approve the use of this system yet? I am eager to have another team use it and give me some feedback. :)

Re: FALCON Open-Source Object Recognition System Now Available!
 

I don't know of a non-JavaScript way of doing that, except I think I read something about DIV tags and CSS being able to do that (it seems like you should be able to set the :hover background image and have it work that way). This approach might not work in Internet Explorer, or at all for that matter.

My mentor approved this system this past weekend. He just needed the info to order it. He ordered the board; now we're just waiting on shipping. It might be in today; I don't know yet (take a peek at this post time. 3:48AM local...).

What exactly do I need to download from Xilinx? I found the WebPack, but it is over a gig and my connection isn't *that* fast...I can get it if I need it, but I have a hard time believing that I need to download a gig of tools to make this work.

Eagerly looking forward to playing with this system,
JBot

EDIT: I figured it out, and I updated the wiki article.

EDIT2: Your teaser pic link is broken. IMHO you should tear it down, but that's up to you.

Re: FALCON Open-Source Object Recognition System Now Available!
 

Thank you!

It's gone now. :)

Re: FALCON Open-Source Object Recognition System Now Available!
 

The other thing I'm seeing that is concerning me is the I/O board in the picture--the board in between the female headers on the Spartan and the camera sensor. Where did it come from?

Just trying to fill in some holes.

JBot

Re: FALCON Open-Source Object Recognition System Now Available!
 

That little adapter is my "cable". :) I had no need for the length of cable for my desktop prototyping board, so I wired up that adapter instead. I don't have a schematic for it, but it shoudn't be too hard to figure it out from the cable construction diagram.

Re: FALCON Open-Source Object Recognition System Now Available!
 

Hey, just wondering how things are going JBot. :)

Re: FALCON Open-Source Object Recognition System Now Available!
 

I have the Spartan board, but I neglected to get a USB JTAG cable (hence the bit of info I put in your wiki regarding the parallel JTAG cable...) and I don't have a parallel port. The new cable is ordered, but I'm still waiting. I haven't had a chance to look at any of the code, so I probably won't get to it before Atlanta. Not to mention I won't even have my laptop until Friday because my sister accidentally took my computer bag with her on vacation...and while I could get the 1.3gb file again and install it here, I don't think I'll bother.

I still might try it, but it will be highly experimental if I get it going...:eek:

I don't know what I'm going to do for a ribbon cable. It would appear that I need a ribbon cable with a male end and a female end--highly rare. I have no idea how I would pull that together either. Do I have a problem with just plugging the sensor directly into the header on the board? I saw that I need to cut a pin; will this make the sensor inoperable with a CMUcam board?

JBot

Re: FALCON Open-Source Object Recognition System Now Available!
 

That's OK; have fun with it in the off-season. :) I am working on getting a demonstration video together; maybe I'll have it in the next few weeks on FirstTube.

You will need to custom assemble that cable according to the instructions on the Wiki. You CANNOT plug the sensor directly into the FPGA board; the power pins (Vcc and Gnd) need to be re-routed through the use of the ribbon cable (or an adapter board).

I have no idea if cutting that pin will make the CMUCam inoperable; it is a redundant +5V power pin that tends to make the FPGA overheat if left connected. You could find and cut the appropriate wire in the ribbon cable instead, I suppose.

I just realized yesterday that I left a critical piece of information off of the wiki! :o

You will need to obtain two 2.2K resistors and solder them between the I2C communications pins and Vcc, otherwise the FPGA cannot configure the module and you will get very interesting pictures using the frame dump option! :)

I will be putting pictures and instructions on the Wiki in the next few days, although it may not be done until after Easter.

Hope this helps, and thanks for being a "ginuea pig", so to speak! :D

Re: FALCON Open-Source Object Recognition System Now Available!
 

I have updated the Wiki with the required information. Thank you for your patience!

I am in the process of getting a demonstration video together; hopefully that will generate more interest in this system. :)

Re: FALCON Open-Source Object Recognition System Now Available!
 

I am going to heat-gun an old system board to swipe a male header off it; I will probably do this when it gets warm outside (it's 40* here in MI) because heat-gunning all that plastic, lead, and fire retardant will most likely make me a bit tipsy if done indoors. I might end up getting a new soldering iron, too--my dad's 12 year old dinosaur probably won't do it.

Thanks for the mental stimulation. Much better than fighting with the CMUcam--now I'll be fighting with the FALCON!

I will post back when I know/do more.

JBot

Re: FALCON Open-Source Object Recognition System Now Available!
 

FALCON Version 1.002 is now available!

Get it at http://www.falconir.com/downloads.php.

Features include:
* Tracking at 9 frames per second, instead of 6
* Protocol change to include the current tracking mode in the tracking string
* Numerous bugfixes and hardening.

Have fun!

Re: FALCON Open-Source Object Recognition System Now Available!
 

And, as promised so long ago, here is a short video of the original 1.001 FALCON system on my Vex 'bot, Loki: http://www.falconir.com/media/falcon...t_vex_demo.mpg

Here is a better picture of the machine, sans light-tracking shades:
http://www.falconir.com/media/falcon...x_demo_bot.jpg

I will be putting together a much higher quality longer demo video in the next couple of months, maybe even on an FRC robot! :)