Stereoacuity and Binocular Correspondence

描述（由申请人提供）：给定3D几何形状，立体视觉对于指定视觉场景的布局可能不如对于提供关于表面形状和结构的局部信息有用。该研究的主要目的是探索立体定义的表面是如何编码的，以及它们在立体匹配和立体灵敏度中的作用。第一个目标将研究编码表面边缘和编码精细表面浮雕的机制之间的相互作用。第二个目标将研究表面编码的连续曲面弯曲的深度。最近的研究表明，视差通常是相对于一个局部的参考框架，即一个相邻的表面，而不是相对于传统的头部为中心或眼睛为中心的坐标系用于表示绝对视差编码。第三个目标将审查的意义，当地的参考框架的背景下，真实的表面和对象。本提案中的实验将检查正常人类观察者的立体声处理。约5%的人口患有威胁正常立体视觉发育的眼功能障碍（斜视）或屈光障碍（屈光参差）。虽然大多数人可以科普立体视觉的丧失，但异常的双眼发育也经常与一只眼睛的敏锐度缺陷（弱视）有关-这是一个更严重的问题。这些对正常观察者的研究将提供关于正常立体视觉的基本知识，这是理解双眼系统异常的先决条件

DESCRIPTION (provided by applicant): Given the 3D geometry, stereopsis is probably less useful for specifying the layout of the visual scene than for supplying local information about surface shape and structure. The major objective of the proposed research is to explore how stereo-defined surfaces are encoded, and their role in stereo matching and stereo sensitivity. The first aim will examine the interaction between mechanisms that encode surface edges and mechanisms that encode fine surface relief. The second aim will examine surface encoding for continuous surfaces curved in depth. Recent studies show that disparity is generally encoded with respect to a local frame-of-reference, i.e. an adjacent surface, rather than with respect to the traditional head-centered or eye-centered coordinate system used to represent absolute disparities. The third aim will examine the significance of local frames-of-reference in the context of real surfaces and objects. The experiments in this proposal will examine stereo processing in normal human observers. About 5% of the population suffers from oculomotor (strabismus) or refractive (anisometropia) disorders that threaten the development of normal stereopsis. While most individuals can cope with the loss of stereopsis, abnormal binocular development is also frequently associated with a deficit in the acuity of one eye (amblyopia) - a more serious problem. These studies of normal observers will provide basic knowledge about normal stereopsis, a prerequisite for understanding abnormalities of the binocular system