Handheld Augmented Reality

Augmented Reality (AR) aims at developing new user interfaces. Although research has produced a large number of application prototypes and AR frameworks in the last 20 years, no project has yet been practical enough to create a mass market success. There are many reasons for this. Traditionally, AR researchers have primarily created prototypes that aim to solve engineering problems such as maintenance or new interfaces for complex environments such as head up displays for navigation and battlefield systems. Most researchers still see AR as a basic research area. Developing easy to use, practical applications, such as for home users, is therefore usually not a goal. Another problem with many Augmented Reality systems is the highly complex hardware setup, often including expensive commercial sensors, input devices and output devices. These devices are often bulky and fragile, since they were never meant to be operated by untrained users. Technology has aimed at moving Augmented Reality to a mass-market. Instead of specialized and expensive hardware, this project targets low cost mobile devices, namely mobile phones. In contrast to traditional AR hardware, people already own these devices and know how to operate them. Recently, processing capabilities of mobile phones have reached a level that makes these devices capable of running standalone AR applications and renders them ideal candidates for mass marketed Augmented Reality solutions. This thesis presents a framework that for the first time allows for the creation of practical AR applications on end user-owned devices. The software runs on a broad range of devices and has been used for several–some even commercial–applications. To prove the applicability of the new platform the author of this thesis has performed evaluations with untrained users in real-life environments such as museum exhibitions or conferences.-iv