Skip to main content
eScholarship
Open Access Publications from the University of California

UC Berkeley

UC Berkeley Electronic Theses and Dissertations bannerUC Berkeley

The Statistics of Disparity in Natural Viewing

Abstract

Humans and many animals have forward-facing eyes providing different views of the environment. Precise depth estimates can be derived from the resulting binocular disparities, but determining which parts of the two retinal images correspond to one another is computationally challenging. To aid the computation, the visual system focuses the search on a small range of disparities. Analysis of the statistics of natural scenes has revealed evidence that the brain exploits regularities to efficiently encode the visual world. We asked whether the brain similarly exploits the disparities encountered during binocular viewing of the natural environment. We did this by simultaneously measuring binocular eye position and 3D scene geometry during natural tasks using a novel device that can reconstruct a participant’s retinal image while they perform natural tasks such as socializing, navigating, and making a sandwich. We find that the natural distribution of disparities is indeed matched to the statistics of natural disparity, allowing for a smaller range of correspondence search. Furthermore, the distribution of disparity explains the perception of some ambiguous stereograms. Finally, disparity preferences of macaque cortical neurons are consistent with the natural distribution. Analysis of the disparities resulting from simulated eye positions yields notably different distributions, suggesting that models of disparity processing should carefully consider how assumptions about eye position could influence the results. We conclude that binocular neural processes are well-matched to the regularities of the 3D environment.

Main Content
For improved accessibility of PDF content, download the file to your device.
Current View