Tozawa Junko, Oyama Tadasu
Kawamura Gakuen Woman's University, Chiba, Japan.
Perception. 2006;35(8):1007-23. doi: 10.1068/p5251.
In three experiments we examined the relative effectiveness of motion parallax and two perspective cues for the perception of size and distance. The experimental stimuli consisted of two ellipses (a standard and a comparison) and a horizontal line that indicated the horizon. The subject's task was to report the apparent size and distance of the comparison stimulus relative to the standard stimulus. Two perspective cues were given by the relative heights of the two stimuli and the absolute height of the standard stimulus below the horizon. Motion parallax was defined by both the ratio and the difference in angular velocities between the two stimuli on the display. In experiment 1 we examined the effects of the two perspective cues and the motion parallax. In experiment 2 we eliminated the horizon line, and examined the role of the horizon in size and distance perception. In experiment 3 we separately evaluated the effects of motion parallax and the relative height cues. The results from the three experiments showed that the effect of motion parallax and the two perspective cues were different in three ways. First, the relative effectiveness of motion parallax and the two perspective cues differed for size and distance estimates. For size estimates, the motion parallax was more effective than the perspective cues (experiments 1 and 3). For distance estimates, the motion parallax was as effective as the two perspective cues (experiments 1 and 3). Second, the role of the horizon differed for size and distance estimates. The size estimates were strongly affected by the horizon, while the distance estimates were not affected much by the horizon (experiment 2). Third, the effective perspective cues differed for the size estimates and the distance estimates: size estimates were affected by the perspective cues as a combination of the horizon and relative height; distance estimates were affected by the perspective cues as an interaction between the absolute and relative heights without the horizon line.
在三项实验中,我们研究了运动视差和两种透视线索在大小和距离感知方面的相对有效性。实验刺激由两个椭圆(一个标准椭圆和一个比较椭圆)以及一条表示地平线的水平线组成。受试者的任务是报告比较刺激相对于标准刺激的表观大小和距离。两种透视线索由两个刺激的相对高度以及标准刺激在地平线以下的绝对高度给出。运动视差由显示器上两个刺激的角速度之比和差值来定义。在实验1中,我们研究了两种透视线索和运动视差的效果。在实验2中,我们去掉了地平线,研究了地平线在大小和距离感知中的作用。在实验3中,我们分别评估了运动视差和相对高度线索的效果。这三项实验的结果表明,运动视差和两种透视线索的效果在三个方面存在差异。首先,运动视差和两种透视线索在大小和距离估计方面的相对有效性不同。对于大小估计,运动视差比透视线索更有效(实验1和3)。对于距离估计,运动视差与两种透视线索的效果相当(实验1和3)。其次,地平线在大小和距离估计中的作用不同。大小估计受到地平线的强烈影响,而距离估计受地平线的影响不大(实验2)。第三,大小估计和距离估计的有效透视线索不同:大小估计受到地平线和相对高度相结合的透视线索的影响;距离估计受到绝对高度和相对高度相互作用且没有地平线的透视线索的影响。