Dynamic View Expansion for Improving Visual Search in Video See-through AR

Abstract

The extension or expansion of human vision is often accomplished with video see-through head mounted displays (HMDs) because of their clarity and ability to modulate background information. However, little is known about how we should control these augmentations, and continuous augmentation can have negative consequences such as distorted motion perception. To address these problems, we propose a dynamic view expansion system that modulates vergence, translation, or scale of video see-through cameras to give users on-demand peripheral vision enhancement. Unlike other methods that modify a user’s direct field of view, we take advantage of ultrawide fisheye lenses to provide access to peripheral information that would not otherwise be available. In a series of experiments testing our prototype in real world search, identification, and matching tasks, we test these expansion methods and evaluate both user performance and subjective measures such as fatigue and simulation sickness. Results show that less head movement is required with dynamic view expansion, but performance varies with application.