運動視差にもとづいた立体動画像から知覚される実対象の奥行きと安定度

抄録

We examined the perception of depth and stability when viewing a video movie made with a camera that moved laterally back and forth 65mm. The two viewing conditions were: (a) head movement synchronized to the stimulus movement and (b) head stationary. In Experiment 1 (N=16), both conditions showed approximately the same magnitude of apparent depth. We argue that (a) in the head movement condition, observers used the cue produced by observer-produced motion parallactic depth, whereas (b) in the no head movement condition, they used the kinetic depth cue or cue(s) from "structure from motion". In Experiment 2 (N=15), the head movement condition showed a greater stability (less perceived motion) compared to the head stationary condition. The results indicate that simulating observer-produced motion parallax (head movement condition) on a video movie of an actual object has no advantages in terms of depth perception. The advantage lies in producing greater stability.