Augmented Reality Reloaded
The Handheld AR project has gone into its second phase. From 2003-2007 we worked on introducing Augmented Reality to mobile phones and PDAs. Now it is time to extend our goals to a new level:
Augmented Reality Anywhere and Anytime
To fullfill this goal we are working on extending our AR platform to run on more devices and to track from no or less obtrusive markers.
The Handheld AR Platform
A standard, off-the-shelf mobile phone constitutes a cost-effective and lightweight hardware platform for Augmented Reality (AR). A mobile phone provides a simple, well-known user interface, and is fully equipped with a camera (and often with a touch-screen too) for providing a video see-through Magic Lens metaphor of interaction. In our Handheld AR framework, all interactive processing is done exclusively on the mobile phone without relying on a server infrastructure. This makes our approach highly scalable to the number of users and number of devices. Because of the low cost and suitable ergonomic properties of the handheld platform, massive multi-user AR applications become possible for the first time.
Test your Windows Mobile phone
We are creating a database of Windows Mobile phone cameras. Check this page to see the current state and help out by using our testing tool on your device.
Technology
Our all new software framework Studierstube ES (StbES) incorporates the experience of 5 years of AR on PDAs and mobile phones. Rewritten from scratch in mid of 2006 is represents a radical new way of doing Augmented Reality on handheld devices. Other than previous attempts it makes no more compromises by using software that was primarily written for other platforms. Studierstube ES was designed from ground up to make best use of small platforms. It is not a prototype, but a a complete solution for AR on mobile phones that can be used for development real live applications of commercial grade.
Studierstube Tracker (StbTracker) is a computer vision library for detection and pose estimation of 2D barcodes. It is a successor to the well known ARToolKitPlus library. Its concept is very similar to that of ARToolKit (ARTK), ARToolKitPlus (ARTK+), ARTag and similar libraries, but its code base is completely different. StbTracker has been written from scratch mid of 2007 with high performance for PCs as well as mobile phones in mind.
StbTracker has been developed from scratch after 4 years of experience with ARToolKitPlus. It was designed from ground up to support mobile phones as well as PCs. Hence, its memory requirements are very low (100KB) and processing is very fast (about twice as fast as ARTK+ on mobile phones). While ARTK+ follows a monolithic approach, requiring configuration at compile time, StbTracker is highly modular: Developers can extend StbTracker in any way, creating new features for it. Different to ARTK+, StbTracker is not open source though.
The Muddleware real-time database is inspired by the concept of a Tuplespace. All data elements are stored as XML nodes, using an extended version of the TinyXML DOM implementation. Clients use XPath to address query or update operations. To reduce the network load all operations can be sent as batches. In addition to classic databases, Muddleware allows clients to register for node updates (Watchdogs). This removes the need for polling and provides a simple, yet powerful publish/subscribe mechanism, which allows easy setup of communication channels between clients. Muddleware uses a memory-mapped database and can handle thousands of queries/sec., which is fast enough to provide real-time performance for hundrets of clients. Muddleware was implemented using ACE and TinyXML and is highly portable.
Even though ARToolKitPlus is not updated anymore, many people are still using it, since it is free of charge and open source under the GPL. ARToolKitPlus has been superseeded by Studierstube Tracker, which implements all features of ARToolKitPlus, but has many more features, better accuracy and is even faster.
Current Projects
Pill Recognition is an application developed within a Master thesis project for online detection, segmentation, measurement and instant recognition of medicals from a mobile phone image. The application connects to a database and provides detailed results of a query using attributes including size, color or shape of the pills detected.
Map Tracking is our latest approach to bring Augmented Reality outside of labs into more natural environments. Instead of classic markers (rectangular or circular) we enhance regular maps with a small grid of dots that cover only ~1% of the map's area. The map tracker performs six degree of freedom pose tracking from uniquely textured planar objects (not just maps) at 15-30 fps overall application performance.
Cows vs. Aliens is a game where players are divided into two competing teams. The game setting is a group of locations which are pastures where cows are quietly grazing; there are also two special locations where a stable for each team is located. Adjacent locations are connected to each other and form a graph. Cows are not aware of the fact that an alien invasion is currently taking place and have no chance to survive without the help from players; UFOs are present on some of the locations and they are ready to shoot any cow that will try to run from that location.
Mobile Augmented Reality Quest (MARQ) aims at developing an electronic tour guide for museums based on a self-contained, inexpensive game consoles, that delivers a fully interactive 3D Augmented Reality (AR) to a group of visitors. The new approach introduces very low costs per device and high scalability which allows for the first time the realization of applications for many concurrent users as required for an appealing museum installation.
All older projects can be found here.
website maintained by
Dieter Schmalstieg
last updated on 2010-05-29
