Suppr超能文献

通过深度进行轮廓补全会干扰立体视敏度。

Contour completion through depth interferes with stereoacuity.

作者信息

Vreven Dawn, McKee Suzanne P, Verghese Preeti

机构信息

Smith-Kettlewell Eye Research Institute, 2318 Fillmore Street, San Francisco, CA 94115, USA.

出版信息

Vision Res. 2002 Aug;42(18):2153-162. doi: 10.1016/s0042-6989(02)00126-8.

Abstract

Local disparity signals must interact in visual cortex to represent boundaries and surfaces of three-dimensional (3D) objects. We investigated how disparity signals interact in 3D contours and in 3D surfaces generated from the contours. We compared flat (single disparity) stimuli with curved (multi-disparity) stimuli. We found no consistent differences in sensitivity to contours vs. surfaces; for equivalent amounts of disparity, however, observers were more sensitive to flat stimuli than curved stimuli. Poor depth sensitivity for curved stimuli cannot be explained by the larger range of disparities present in the curved surface, nor by disparity averaging, nor by poor sensitivity to the largest disparity in the stimulus. Surprisingly, sensitivity to surfaces curved in depth was improved by removing portions of the surface and thus removing disparity information. Stimulus configuration had a profound effect on stereo thresholds that cannot be accounted for by disparity-energy models of V1 processing. We suggest that higher-level 3D contour or 3D shape mechanisms are involved.

摘要

局部视差信号必须在视觉皮层中相互作用,以表征三维(3D)物体的边界和表面。我们研究了视差信号在3D轮廓以及由轮廓生成的3D表面中是如何相互作用的。我们将平面(单视差)刺激与曲面(多视差)刺激进行了比较。我们发现对轮廓和表面的敏感度没有一致的差异;然而,对于等量的视差,观察者对平面刺激比曲面刺激更敏感。曲面刺激的深度敏感度较差,这既不能用曲面中存在的更大视差范围来解释,也不能用视差平均来解释,也不能用对刺激中最大视差的敏感度差来解释。令人惊讶的是,通过去除部分表面从而去除视差信息,对深度弯曲表面的敏感度得到了提高。刺激配置对视差阈值有深远影响,这无法用V1处理的视差能量模型来解释。我们认为涉及更高层次的3D轮廓或3D形状机制。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验