Abstract
Judging the height and laterality of an object requires vertical and horizontal references that coincide respectively, with the judged eye level and midline of the body trunk. What visual cues determine these references? We previously found that the focus of expansion (FOE) of the optic flow determines judged eye level (Wu et al, VSS02). Here, we extended the study to the horizontal reference. In a virtual environment, observers judged the midline while viewing optical flow display with its FOE shifted horizontally (±10, ±20 and ±30deg). As predicted, horizontal midline was biased to the FOE [rate=0.20 (left); 0.29 (right)]. Arguably, besides showing the utility of the optic flow, our study also implies that other stable cues, eg linear perspective cue (Sedgwick, 1986), can help determine the references. To test this, observers judged their eye levels in the dark. In condition 1, two fluorescent lines (2.5m long) were placed on the right wall with their relative orientation varied to cause false linear perspective. Compared to baseline (parallel lines/true linear perspective), judged eye level was affected by false linear perspective. For example, with one line parallel to the floor, judged eye level changed with the orientation of the second line. If the two lines converged below the actual eye level, judged eye level shifted down. That the effect was mainly due to false linear perspective was shown by removing the line parallel to the floor; the orientation of the second line produced only about 1/9th of the effect. In condition 2, we placed a fluorescent line on each side of a wall (1.5m apart). While the pair of lines remained parallel at the same height, their orientation relative to the floor varied (0, 5, 10deg). Again, relative to baseline, judged eye level changed at a rate of 0.18 as the orientation of the lines varied, compared to a single tilted line (rate=0.06). These findings highlight the role of environmental constants in space vision.