Purchase this article with an account.
Nick M. Barnes, Janine G. Walker, Chris McCarthy, Viorica Botea, Adele F. Scott, Hugh Dennet, Paulette Lieby; Evaluating Depth-based Visual Representations For Mobility In Simulated Prosthetic Vision. Invest. Ophthalmol. Vis. Sci. 2012;53(14):5550.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Safe and efficient mobility is an important challenge for prosthetic vision devices. To date, research into possible visual representations has largely employed intensity approaches that represent luminance. However, depth-based representations may provide advantages: robustness to low contrast; and, perception of distance to unfamiliar overhanging obstacles from a single view. We investigate whether a depth-based representation for mobility has advantages compared to intensity-based representations in a controlled high-contrast environment; a setting beneficial to intensity-based visual representations.
11 normally-sighted (20/20, Pelli-Robson>=1.95) participants traversed a controlled high-contrast indoor mobility course using a mobile prosthetic vision simulator set with phosphenes centrally displayed identically to both eyes. We compared a depth-based representation (Depth; 35x30 phosphenes, 16 separately noticeable levels of phosphene brightness) to two intensity-based representations: high dynamic range Intensity (35x30, 64 levels); and, Degraded (6x4, 2 levels). The study had a double-blind randomized factorial
In the test phase, mean PPWS was more than 43.5% for Intensity and Depth. Across all sessions, Intensity (n=335) had significantly higher PPWS than Depth (n=331; p<0.001). Both Depth and Intensity were faster than Degraded (n =55; 20.25% PPWS for Degraded, p<0.001). There was an interaction effect for VR and Obstacle Placement (p=0.005). Specifically, for Intensity, PPWS reduced significantly with the introduction of obstacles (p<0.001); whereas, PPWS did not significantly change for Depth. PPWS improved significantly over the sessions (p<0.001) for Depth and Intensity.
Both Depth and Intensity were effective for navigation. For Depth, participants were not significantly slowed by overhanging obstacles, while they were for Intensity, despite high contrast and greater dynamic range. Depth shows promise as an effective approach for mobility for prosthetic vision devices.
This PDF is available to Subscribers Only