Last year’s Android Developer Challenge brought us such augmented reality applications as Wikitude and Enkin (which, until I’m proven wrong, is the first AR vaporware).
This time around, we are likely to see a surge in the number of AR applications participating in the challenge. An early bird, geoPaste, is coming to us all the way from Australia.
As can be seen in the video above, geoPaste lets you annotate reality right through your mobile phone, by sketching little pictures (or loading previously created images). Developer Gary Silva is concentrating at the moment on winning the challenge, but he has a grand vision for his app, “to make AR publishing available to everybody”.
After ADC1 I had also seen a Wikitude demo and saw their slant towards providing AR content for consumers, like Enkin ( and which Layar now continues ). From there I thought that AR publishing could be put directly into the hands of end users and was inspired to start working on geoPaste to try to realize that vision. My original thoughts were along the lines of digital graffiti, legal and harmless but at the same time indelible, hence my drawings metaphor.
Good luck on the challenge!
More details, over here.
Filed under: Uncategorized | Tagged: geoPaste, GPS and Compass, Handhelds and Cellphones, Tourism and Outdoors |
Games Alfresco 
I loved the demo where you just point to a location in your field of view and the drawing is mapped to that location.
Then I saw step 2: positioning the drawing on a map – which threw me off. Isn’t that a repeat of the same action ?
(I realize it’s to compensate for low accuracy, but that feels like an excuse from a user perspective…)
Also why do you need an AR button while the phone can recognize that you are holding it up in “AR mode”?
But seriously – great initiative and keep it up!
Hi Ori Inbar,
Yes it does seems to be a repeat of the same action and is intended to compensate for local GPS inaccuracy.
However, in the current version of the application it also allows you to publish geoPaste drawings anywhere in the world, remotely and not just within your immediate vicinity.
Regards
Wow, that’s cool. Course maybe if they used basic image recognition, along side the compass, posted drawings could be better locked to their real world locations. =/
They’ve sort of beat metaio World to it’s *albeit rough* goal.
I love this, as ugc seems to be a secondary thought to AR worlds at the moment.
—
I think to have really accurate mapping there will need to be some pretty big steps from major players.
Specifically, I think we might need point clouds of citys that AR software can use to align to their environment.
Point clouds generated and generated, preferably, from the mass’s of photographs already online. Unfortuntely Microsoft seems to just want to keep Photosynth a pretty method to share photos rather then putting it towards anything so grand :-/
Accurate registration is a holy grail for AR whatever the methods employed, from classic localized marker based systems and barcodes through compass/gps/gyro assisted, image recognition against pre-existing photos and point clouds etc. Each have benefits, deficiencies and computational overhead (sometimes massive) particularly for a real time experience on cell phones. My current bias is to towards compass & gps, I’m looking forward to more & more accurate sensors and systems in these areas being incorporated into portable devices. I think newer non military gps systems (particular multiple gps systems used together Galileo & gps etc ) hold great promise to achieve this in the years to come.
Galileo will definitely improve things.
I think when hybreded with gps we might get down to a the meter or even half-meter range.
Which is good enough for hovering stuff, or even free-standing AR stuff.
So we should have that in 4 years time.
However, to actualy “skin” existing stuff. (AR graffiti on a wall), I think having existing 3d models of citys to align too is the best longterm option. That should give us cm or even mm accuracy.
Of course, its a hhhhuuuugggeee amount of work, but I think if its possibly for a community to do it themselves, ver photographs, its a task that could be accomplished.