Vertical gaze angle: absolute height-in-scene information for the programming of prehension

Citation
Pl. Gardner et M. Mon-williams, Vertical gaze angle: absolute height-in-scene information for the programming of prehension, EXP BRAIN R, 136(3), 2001, pp. 379-385
Citations number
24
Categorie Soggetti
Neurosciences & Behavoir
Journal title
EXPERIMENTAL BRAIN RESEARCH
ISSN journal
00144819 → ACNP
Volume
136
Issue
3
Year of publication
2001
Pages
379 - 385
Database
ISI
SICI code
0014-4819(200102)136:3<379:VGAAHI>2.0.ZU;2-C
Abstract
One possible source of information regarding the distance of a fixated targ et is provided by the height of the object within the visual scene. It is a ccepted that this cue can provide ordinal information, but generally it has been assumed that the nervous system cannot extract "absolute" information from height-in-scene. In order to use height-in-scene, the nervous system would need to be sensitive to ocular position with respect to the head and to head orientation with respect to the shoulders (i.e. vertical gaze angle or VGA). We used a perturbation technique to establish whether the nervous system uses vertical gaze angle as a distance cue. Vertical gaze angle was perturbed using ophthalmic prisms with the base oriented either up or down . In experiment 1, participants were required to carry out an open-loop poi nting task whilst wearing. (1) no prisms; (2) a base-up prism; or (3) a bas e-down prism. In experiment 2, the participants reached to grasp an object under closed-loop viewing conditions whilst wearing: (1) no prisms; (2) a b ase-up prism; or (3) a base-down prism. Experiment 1 and 2 provided clear e vidence that the human nervous system uses vertical gaze angle as a distanc e cue. It was found that the weighting attached to VGA decreased with incre asing target distance. The weighting attached to VGA was also affected by t he discrepancy between the height of the target, as specified by all other distance cues, and the height indicated by the initial estimate of the posi tion of the supporting surface. We conclude by considering the use of heigh t-in-scene information in the perception of surface slant and highlight som e of the complexities that must be involved in the computation of environme ntal layout.