Summary for week 3
The Argon AR Web Browser and Standards-based AR Application Environment
Augmented Reality (AR) has been developed in 1965 by Ivan Sutherland, but researches in this field only began in the late 1980s. Now, this technology should evolve to AR desktops gathering independent applications in a single environment. That’s why Argon, an AR web browser, is made for, it gathers independent channels in a virtual environment, each channel having its own user interface and content.
The motivations for this project which began in Fall 2009 have been that technologies on mobile phones are now developed enough, and also that mobile browsers are similar to computer browsers and that Apple allows 3D rendering of 2D contents in Safari. Although mobile phones are not the most immersive device, it is widespread and the immersion could be improved with a HMD. The first goal is to broadcast AR systems in order for them to become available for everyone, so that this technology can be studied in depth. The second aim is to use an existing mobile technology. The last goal is to build a system allowing to create both simple and elaborated applications. Argon uses web standards and KML, as well as KARML, an extension of KML for AR.
Tools for creating AR systems already exist, but AR browsers emerged recently. Also, an AR extension to KML (ARML) has already been developed but provided only some additional features. So Argon represents the next generation of tools for authoring AR systems. Some tools can be used along with Argon, particularly Google Street View, Photosynth and Bing Maps which allow the user to create and visualize 3D panoramas. These panoramas can be imported into Argon using KARML, for live or remote use.
For what concerns KARML, its aim is to bind a content and its physical location. KML is used because it is widespread, and because HTML 2.0 and COLLADA 3D bindings have already been developed. KARML is more efficient than KML because in addition to giving the position of a content, its scaling and orientation can be modified. Furthermore, KARML allows relative and absolute location whereas KML only allows absolute location.
However, one issue of this project is that the embedded GPS sensors have only a 10 meter accuracy, leading to many vaguenesses. A solution has been to allow the user to override the location with another close location called GeoSpot. Moreover, each user can improve the precision of a GeoSpot. A future enhancement will be to give to a user located at a GeoSpot the ability to replace the video with panoramas having the same orientation as the device.
Finally, 7 examples of projects having been developed inside or outside Georgia Tech show the capabilities of Argon, for instance rapid development or blending 2D interfaces and AR.
Although Argon is already an efficient tool, several enhancements can be made and will be developed in upcoming months and years. Among the most important improvements, Argon will support full 3D model rendering, and will also support other markup languages. Another great enhancement will be for Argon to be supported on desktop and other mobile phones.
Thus, Argon is a powerful AR web browser and should allow a dramatical expansion of AR. But, in order for users to really be interested in AR and use Argon on a daily basis, news could be integrated (if it has not been done yet). So, could an improvement be to retrieve local news and to display them in Argon?
Comments are closed.