Augmented reality type apps are what all the cool geeks are doing it seems. I mean the apps that use a video camera to recognise certain symbols (aka fiducials) and then overlay additional virtual content. Like this prototype I built.
I think there’s real space for interesting apps here. The technology has been getting better and more accesible for a while. One remaining constraint though is that the markers came in very computery designs and there wasn’t much you could do to get around that. The software systems depended on very formula-based symbols, which might be great for computer vision but didn’t leave much aesthetic control with the creator.
Things are changing for the better though on that front too. As this article explains the d-touch marker recognition system is now much more flexible in terms of the markers it will recognise. For example the speech bubble image here is recognisable to the system, a big improvement from the more typical amoeba style blob shown above it.
I’m going to play with some prototype ideas I have for using this new system, so hopefully will be able to kick out some stuff soon.
Geeky d-touch setup details
I had a lot of trouble finding documentation on how to get the DTServer application running, so here’s a quick run through of what worked for me (using Processing on Mac OSX Leopard)
- grab DTServer and Processing Client from SourceForge
- download d-touch Analyser
- create a marker of your own following the design rules on the d-touch site
- run the marker through d-touch Analyser to make sure it is valid
- use the Copy ID button in d-touch Analyser to grab the ID of the marker (e.g. 0, 1, 1, 7, b)
- copy the file seq.txt from DTServer DMG file into Applications folder where you dropped DTServer.app
- edit seq.txt and add a new line with the ID for the marker your just created
- launch DTServer.app, twiddle the video settings (or don’t) and OK to get the server running
- if all went well DTServer should recognise and highlight the marker when you hold it in front of your camera (assuming good lighting, focus, not too far from the camera…)
- run the Processing client sketch dt_demo and you should see a display that reflects the position and rotation of the marker in front of the camera
- that’s it, about all I know on this subject. I only played with this for a couple of hours and I’m definitely not an expert on the d-touch system. I just thought that my experience might be useful to someone else trying to figure this out.