Abstract
Abstract: :
Purpose: was to test subjects' accuracy for locating landmarks using a coupling between their head movements and the resulting optical transformations produced in a computer generated scenery. Methods: Stimulus trials either simulated locomotion across a tree-filled plane while looking at a particular tree (dynamic conditions) or corresponded to a fixed viewpoint (static conditions). As the coupling introduced a lateral motion parallax effect recognized as a determinant cue for distance perception, we expected that subjects would use it for better localization. Sixteen subjects were shown short motion sequences, representing a fixed line of three simplified trees (FT) plus two additional trees (AT) with varying positions to be reproduced. Three conditions (with coupling, without coupling, coupling's replay) were used. After viewing each static or dynamic sequence in perspective, subjects were asked to replace both ATs on a plan view. Results: In static conditions spatial localization was effectively improved. In dynamic conditions however, the better performances were shown in couplings' replay rather than in coupling itself. Conclusion: We may assume that a proper use of artificial lateral motion parallax would need more training for a moving observer than for a static one. Supported by Swiss FNRS Grant 1114-057839.99
Keywords: 477 motion-2D • 499 optic flow • 583 space and scene perception