Abstract
Angular declination (AD) is a reliable cue to egocentric distance when targets are floor-level and within the intermediate distance range; elevated objects can be localized based on where their supporting structures contact the ground. In the real world, humans must locate objects supported by surfaces (tables, countertops, etc.) when the ground contact is obscured or not salient. Can knowledge of height above the ground be used in conjunction with target AD when the ground contact of the support is not visible? Here, observers walked without vision to previewed targets that were vertically-oriented boards resting on the ground, 71 cm high and 2.5-5m distant. One end was yellow; the rest was color-matched to the blue-gray carpet. Targets were oriented with the yellow end up or down, and the viewing duration was either short (< 20 ms, sufficient to detect only the yellow end of the object) or long (100 ms, sufficient to detect both ends). Thus, in the up position, the short viewing duration prohibited detection of the ground contact. Observers were informed the object extended to the height of a typical table and were shown real-world examples in advance. Response sensitivity was similarly high in all conditions (slopes near 1). In Experiment 1, the viewing duration was always short, and underestimation was greater in the yellow-down condition (-17%) relative to yellow-up (-1%). In Experiment 2, the targets were always yellow end up, and underestimation was greater when viewing was long (-24%) relative to short (-14%). Several accounts for the bias differences will be considered, but there was no indication that a visible ground contact is crucial. The findings indicate an ability to utilize knowledge of object height to localize objects, but the change in bias across viewing durations suggests a preference to use the ground contact when it becomes visible.
Meeting abstract presented at VSS 2015