Abstract
We developed the DSpecs as a wearable visual aid utilizing an augmented reality (AR) platform, eye tracking system, a portable mini PC, and a battery. Video processing algorithms of a scene camera in the DSpecs produced personalized visual augmentation profiles to functionally expand their limited PVF. Processed videos were displayed in real time on the AR display based on prior binocular field measurements obtained using the DSpecs diagnostic mode. Twenty patients walked with the DSpecs in a validated obstacle track once with visual augmentation being activated (with expansion) and another time without it (without expansion). The two walking conditions were repeated for each patient and gaze data were recorded and averaged. We calculated four parameters to describe the eye scanning behavior while walking. The parameters were: number of fixations per second, variance of fixation locations, central gaze location score, and mean saccadic amplitude.
