Usability Engineering of Text Drawing Styles in Augmented Reality User Interfaces

In the coming years, augmented reality, mobile computing, and related technologies have the potential to completely redefine how we interact with and use computers. No longer will we be bound to desktops and laptops, nor will we be bound to monitors, two-dimensional (2D) screens, and graphical user interface (GUI) backgrounds. Instead we will employ wearable systems to move about and augmented reality displays to overlay 2D and three-dimensional (3D) graphics onto the real world.

When the computer graphics and user interface communities evolved from text-based user interfaces to 2D GUIs, many in the field noted the need for “new analyses and metrics“ [Shneiderman et al., 1995]; the same is equally true today as we shift from 2D GUI-based user interfaces and environments, to 3D, stereoscopic virtual (VR) and augmented reality (AR) environments. As we rush to advance the state of technology of AR and its capabilities, we need to advance the processes by which these environments are designed, built, and evaluated. Along these lines, this dissertation provides insight into the processes and products of AR usability evaluation.

Despite the fact that this technology fundamentally changes the way we visualize, use, and interact with information, very little HCI work in general, and user-centered design and evaluation in particular, have been done to date specifically in AR [Swan & Gabbard, 2005]. While traditional HCI methods can be successfully applied in AR to determine what information should be presented to the user [Gabbard, 2002], these approaches do not tell us, and what, to date, has not been researched, is how information should be presented to the user.

A difficulty in producing effective AR user interfaces (UIs) in outdoor AR settings lies in the wide range of environmental conditions that may be present, and specifically large-scale fluctuations in natural lighting and wide variations in likely backgrounds or objects in the scene. In many cases, a carefully designed AR user interface may be easily legible under some lighting and background conditions, and minutes later be totally illegible in others. Since lighting and background conditions may vary from minute to minute in dynamic AR usage contexts, there is a need for basic research to understand the relationship between real-world backgrounds and objects and associated augmenting text drawing styles.

This research identifies characteristics of AR text drawing styles that affect legibility on common real-world backgrounds. We present the concept of active text drawing styles that adapt in real-time to changes in the real-world backgrounds. We also present lessons learned on applying traditional usability engineering techniques to outdoor AR application development and propose a modified usability engineering process to support user interface design of novel technologies such as AR.

Results of this research provide the following scientific contributions to the field of AR:

Empirical evidence regarding effectiveness of various text drawing styles in affording legibility to outdoor AR users.

Empirical evidence that real-world backgrounds have an effect on the legibility of text drawing styles.

Guidelines to aid AR user interface designers in choosing among various text drawing styles and characteristics of drawing styles produced by the pilot and user-based studies described in this dissertation.

Candidate drawing style algorithms to support an active, real-time, AR display system, where sensors interpret real-world backgrounds to determine appropriate values for display drawing style characteristics.

Lessons learned on applying traditional usability engineering processes to outdoor AR.

A modified usability engineering process to assist developers in identifying effective UI designs vis-à-vis user-based studies.