Foundations of spatial vision: From retinal images to perceived shapes
Foundations of spatial vision: From retinal images to perceived shapes
复制标题
DOI：
10.1037/0033-295x.107.1.6
复制
发表时间：
2000-01-01
影响因子：
5.4
通讯作者：
Craft, WD
中科院分区：
文献类型：
--
作者：
Lappin, JS;Craft, WD
视觉是基于物理上不同的结构之间的空间对应关系--环境、视网膜、大脑和知觉。对环境表面和它们的视网膜图像之间的对应关系的检查表明，这由与局部表面形状相关的2维2阶差分结构(实际上是4阶差分结构)组成，这表明这可能是空间信息的原始形式。接下来，对用于检测独立图像特征之间的相对运动和双眼视差的超锐度进行的实验表明，空间位置在视觉上由周围的光学图案指定，而不是由视网膜坐标指定，受三维物体运动产生的随机图像扰动的影响最小。因此，视网膜图像空间包含四阶微分结构。这种原始空间结构构成了关于局部表面形状的信息。
Vision is based on spatial correspondences between physically different structures-in environment, retina, brain, and perception. An examination of the correspondence between environmental surfaces and their retinal images showed that this consists of 2-dimensional 2nd-order differential structure (effectively 4th-order) associated with local surface shape, suggesting that this might be a primitive form of spatial information. Next, experiments on hyperacuities for detecting relative motion and binocular disparity among separated image features showed that spatial positions are visually specified by the surrounding optical pattern rather than by retinal coordinates, minimally affected by random image perturbations produced by 3-D object motions. Retinal image space, therefore, involves 4th-order differential structure. This primitive spatial structure constitutes information about local surface shape.