Abstract
Reliably judging absolute distances of objects beyond the near distance range (>2 m) is important for planning and guiding actions, such as navigation. Yet, the underlying binocular mechanisms are not fully understood. We thus investigated human observers’ ability to discriminate the absolute distances of two targets that were presented consecutively in the real environment, using a two-interval forced choice (2IFC) procedure with 1 sec inter-stimulus-interval. We also investigated the effect of target duration (0.10, 0.15, 0.25 and 2.0 sec). Paired, 0.2 degree dimly-lit (0.16 cd/m2) targets in a dark room were presented from an array of seven targets located over a range of 4.5-7.0 m from the observer and elevated 0.15-0.75 m from the floor. All targets had the same angular declination (12.5 degree), except the nearest one (11.9 degree), when viewed from an eye-level of 1.7 m. The relative binocular disparities between the farthest target and other targets were 3.13, 6.25 and 16.08 arc min. The targets were presented together with a 2x6 parallel array of dimly-lit (0.04 cd/m2) elements on the floor (1.17-7.0 m), which constituted a texture background in the otherwise dark room. The observers’ task was to report the 2IFC interval with the nearer target. The average results (n=4) revealed the percentage correct in judged depth increased significantly with relative binocular disparity between the two targets, as well as with target duration (p<0.005). We also conducted two control experiments: (i) binocular-dark (without texture background); (ii) monocular-texture background (viewed with motor dominant eye). We found even with a 2-sec target duration, the average performance was below the 75% discrimination threshold. Overall, our results indicate binocular vision and the texture background, acting as a reference frame, play a critical role in reliably judging absolute distances of targets above the ground. Furthermore, the binocular visual system is quite efficient.