Displaying Readable Text in a Head-Tracked, Stereoscopic Virtual Environment

Abstract

In a head-tracked, stereoscopic virtual environment, many straightforward text implementations suffer from poor readability or unnatural behavior. For example, scan-converted text often appears blurry or "shimmery" due to rapidly alternating text thickness because scan conversion depends on the user's location and the user rarely stays perfectly still. Likewise, bitmapped fonts cannot generally mimic objects with fixed size and location because they do not scale and thus do not appear larger as the viewer moves closer. This paper describes a simple method for displaying readable text that need not have a fixed location in the virtual environment, such as menu-system and annotation text. Our approach positions text relative to the user's view frustums (one frustum per eye), adjusting the 3D placement of each piece of text as the user moves, so the text occupies a constant location in each of the view frustums and projects to the same pixels regardless of the user's location. The result is crisp, clear text, consistently fused stereo vision, and reduced visual fatigue compared to many other types of text in virtual-reality environments.