Purchase this article with an account.
Gislin Dagnelie, Greg Seifert, Paul Gibson, Arup Roy, Michael P Barry, Avi Caspi; Retinal prosthesis wearers benefit from information reduction by distance-selective filtering. Invest. Ophthalmol. Vis. Sci. 2017;58(8):4686.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Wearers of retinal prosthesis systems have trouble detecting relevant visual information in complex scenes. Irrelevant information often originates in different depth planes, and could thus be separated and suppressed. In this study we tested the ability of a prototype stereo camera and depth filtering unit to assist Argus II wearers in detecting and discriminating objects.
We developed a stereo camera pair with 14.0 cm base disparity, ¼” VGA (640x480) CMOS sensors, and 2.84 mm focal length, providing 7 arcmin resolution. The cameras, mounted on a flat rigid frame, are clipped onto the Argus II glasses. A USB 3.0 connection transfers imagery to a notebook computer running disparity computation software at 10 frames/s. Two controls, filter center distance and depth bandwidth, allow the user to set the range of depths to be rendered . A video stream to the Argus II video processing unit (VPU) replaces the standard camera input. Performance tests included operations at far (1 or 2 chairs at 1–3 m) and close (canister, cup, or bag of candy at 0.3–0.7 m) ranges, and in either case consisted of 3 tasks: presence (Y/N, 2-AFC, 10 trials), lateral position (L/R/B, 3-AFC, 9 trials), and depth discrimination (R closer/farther than L, 2-AFC, 6 trials, depth separation varied to estimate thresholds), with conditions presented in random order. The presence and position tests were performed with and without distance filtering.
All 4 Argus II users were 100% correct on the chair presence test with filtering; without filtering 3 performed at chance, while the 4th scored 80% (n.s.) but took 49 s (17 s with filtering). On the near presence test subjects were 90-100% correct with, vs. at chance without, filtering. One subject was able to perform the chair position test both with and without filtering, while another was at chance in both conditions; for the near position test all subjects were above chance (67–89%) with, and at or near chance (22–56%) without filtering. For the depth discrimination at 1.50 m, the minimum depth difference successfully detected was 10–20 cm, while at 40 cm a depth difference of 2.5 cm was reliably detected.
Argus II users can benefit from depth-filtered imagery in 2 ways: 1) by detecting the presence of objects within a range of interest; 2) by estimating relative distances of objects to within 10%. A production version of the system is under development.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.
This PDF is available to Subscribers Only