Idle developers of the world take inspiration from this project and unite to create your own version. It’s called the Junkyard Jumbotron because it takes many different displays and allows them to be used as one big interactive display. The image above shows a collection of smartphones displaying a test pattern. The pattern is unique for each device and is used to calibrate the display. Using a digital camera, a picture of these test patterns is snapped, then sent to the server. The server calculates the position of each of the screens, then sends the correct slice of a large image back to each phone.
It’s funny that they use the word Junkyard in the name of the software. Each display needs to be able to run a web browser so you can’t just use junk displays. But one nice side effect of the hardware requirements is that you can still do things like panning and zooming as seen in the video after the break. Here’s the real question: can you make this work as an open source project? How about something that can be easily set up to work with a LAMP server?
[vimeo=http://vimeo.com/20962561 w=470]
[via Reddit]
Try Synergy.
Open source. Cross platform.
http://synergy-foss.org/
Lets you control your cursor across multiple displays. Not exactly the same functionality but in someways better.
http://civic.mit.edu/blog/csik/junkyard-jumbotron says it’s “free and open.” I think my idea of free and open is different from theirs.
It also has code to “map out” (albeit manually) screen locations that may be useful.
VNC can also allow you to do similar things in terms of remote viewing of desktops but full video probably will not work.
You can also achieve this using linux or windows using multiple video cards in a PCI-e machine (or even PCI) and VLC has (had?) provisions to break up a huge video for display on “jumbotron” displays.
Yowza that’s painfully slow. Neat idea though.
http://grok.lsu.edu/article.aspx?articleid=14623
VLC has the “wall video filter” module and is cross platform and open source as well.
The Junkyard Jumbotron is already an open source project. Code is on Github at https://github.com/c4fcm/Junkyard-Jumbotron
We’re very interested in building an open source team to push it forward.
Very cool idea and application! Thanks for the creative input.
It wouldnt be hard to build, if anyone is instrested they can contact me to request an open source development on this which would be compatable with all servers. P.S you wouldnt need to take a picture of the position of the screen. It would just require Javascript, PHP and MySQL.
VLC? Synergy? Obviously you don’t get the point. Read it again.
does it use openCV?
yah I know right cause VLC, Synergy and X11 are actually useful… unlike this. Cross device as long as it has a browser blah blah blah.. its still soo slow that its useless for anything other than a tech demo and it certainly can’t sync sound and video :/ which is the sort of cool stuff you would want on a jumbotron to begin with.
The trick is in the web-browser.
Each device gets a unique id number, then the location of those id’s are pulled from the picture that gets sent in. The location is set as the center of the picture in the page sent to the device with that id number. A little ajax style action allows the pages to be updated with changes generated by the other pages. The picture itself is never split up, but is sent whole to each device with a different centering.
It all came to me when I noticed that the examples were all still pictures and not video, though if devices supported html5 I imagine the same trick would work if the clocks were synced up.
Used three different browser windows on my laptop screen and did a print screen of it. Sent that in and it worked peachy. (Except the MS IE piece of cra6 doesn’t scale right, but fixed it.)
VNC and multi video cards do it, sure. The brilliance here is the analysis and calibration based on an image of the monitors – this IS rather brilliant.
Hey Folks,
There seems to be a lot of interest in the Junkyard Jumbotron here. I want to repeat that I really would like to see an open source development community form around this project. Like others, I really want to see this web based, visual calibration approach work with video, games, etc. If anyone wants to help with this, I’ve posted the url for the source on github. Also, please get in touch with me at jumbotron at media dot mit dot edu.
I am thinking similiar to Max, but instead of using Ajax I would go with Comet. Each unit could go to the same page but with a different ID, and the server side Comet pusher would know throw out the partial images simultaneously, but different chunks based on the id of the unit supplied
Looks like someone who worked with the original team branched the code and may have gotten video to work: https://github.com/JeremyRubin/Junkyard-Jumbotron
Can anyone verify, and maybe post this somewhere so the washed up coders (like me) could get a look and see if it does in fact work?
Thanks,