Spatial and temporal characteristics of perceptual completion at the blind spot

盲点的知觉补全是视觉系统恢复外部世界的重要功能之一，其中盲点的知觉补全代表了物理空间的恢复。以往对盲点知觉完成的研究主要是定性的，而不是定量的（如Ramachandran， 1992）。在本研究中，我们通过使用带间隙的线刺激来定量地检查盲点处的知觉完成。测试线横过盲点，使间隙正好落在观察者右眼的盲点上。观察者被要求报告，测试线似乎与参考线相同，参考线是连续的，呈现在颞视网膜上，与测试线的偏心相同。感知补全定义为测试线的间隙出现与参考线一样的填充，并测量发生填充的线段的最小长度。通过六次实验得到如下结果：(1)当盲区相对两侧的线段很短时，它们表现为分别位于不同位置的两条线段。(2)直线垂直方向上可发生填充的线段最小长度大于水平方向上的线段最小长度。因此，在盲点处存在知觉完成的各向异性。(3)感知补全的各向异性不依赖于盲点横轴和纵轴的差异。(4)测试线的宽度和曝光时间对盲点感知完成度有影响。(5)我们发现的各向异性在另一种刺激模式（如线光栅）中得到证实。(6)盲点处的感知补全容忍线段的不对准、方向和对比度的差异。

The present study examined perceptual completion, or filling-in, at the blind spot Perceptual completion is one of important functions of the visual system in order to restore the outer world Specifically, perceptual completion at the blind spot is representative of the restoration of physical space. Previous studies on perceptual completion at the blind spot were mainly qualitative, not quantitative (e.g., Ramachandran, 1992).In this study, we examined quantitatively perceptual completion at the blind spot by using a line stimulus with a gap. The test line was presented across the blind spot so that the gap fell exactly on the blind spot of observer's right eye. Observers were required to report that the test line appeared to be identical to the reference line, which was continuous and presented to the temporal retina with the same eccentricity of the test line. Perceptual completion was defined as that the gap of the test line appeared filled in as like as the reference line and the minimum length of the line segment for filling-in to occur was measured. The following results were obtained from the six experiments.(1)When the line segments in the opposite sides of the blind spot were very short, they appeared as the two line segments located separately in different positions.(2)The minimum length of the line segments for filling-in to occur was larger in the vertical orientation of the line than that in the horizontal orientation. Thus, there is anisotropy of perceptual completion at the blind spot.(3)The anisotropy of perceptual completion does not depend on the difference of the horizontal and vertical axis of the blind spot.(4)The width and exposure duration of the test line affected on perceptual completion at the blind spot.(5)The anisotropy we found was confirmed by another stimulus pattern, such as line grating.(6)Perceptual completion at the blind spot tolerated differences in misalignment, orientation, and contrast of the line segments.