Purchase this article with an account.
Chien-Chung Chen, Lok-Ting Sio; The interaction between visual mechanisms for 2D and 3D information processing. Journal of Vision 2012;12(9):1041. doi: 10.1167/12.9.1041.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
In a nature environment, a three-dimensional (3D) object often has a 2D pattern on its surfaces. Thus the visual system often processes 2D and 3D information simultaneously. Here, we showed that the 3D structure of an object affects the perception of 2D patterns and the brain regions for these two types of information inhibit each other. The 2D patterns were symmetric patterns consisted of random dots. The symmetric patterns had one half being the reflection of the other half about a central vertical axis. The 3D structures were created by the disparity between the left and the right eye images. The coherence threshold for symmetry detection was measured (2AFC, PSI staircase) with nine types of 3D structures. Compared with the frontoparallel condition, the threshold increased whenever the two sides of symmetric axis were not coplanar. The thresholds for the slant and the flat conditions were similar. Thus, it is coplanarity, rather than depth per se that was important for the 3D effect. BOLD activation was measured on a Bruker 3T scanner (TR=3s) with a multiple block design (18s per block) in which the stimuli varied either in 3D structure, 2D coherence or both. A ventral occipital area showed greater activation to symmetric patterns than to random dots but the difference decreased when there is a depth contrast in the image. The areas V3A/B showed greater activation to the depth than to the flat conditions. This difference was greater with the random dot 2D patterns than with the symmetric ones. Our results showed that while there is a clear functional segregation between the symmetry area in the ventral stream and the depth area in the dorsal stream, there is much inhibitory interaction between them. Both psychophysics and fMRI results show that 2D and 3D information are not processed independently.
Meeting abstract presented at VSS 2012
This PDF is available to Subscribers Only