Texture segmentation without human V4

Norman, L.J.; Heywood, C.A.; Kentridge, R.W.

Abstract

Texture segmentation, or second-order segmentation, is a rapid perceptual process, allowing object and surface boundaries to be effortlessly detected. It is currently unclear whether this is achieved in early cortical areas or whether it necessitates the region referred to as human V4. The present report presents a single case study of patient MS, whose bilateral occipitotemporal damage includes the putative human V4 area, yet whose early visual cortex is spared. As shown in these experiments, MS can accurately locate a target defined by an orientation contrast to its background, even with considerable orientation noise. Importantly, his performance was significantly reduced when the texture edges were masked by black borders (thus preventing edge-based segmentation), indicating that he retains a functional edge-based texture segmentation process. Additionally, when the sign of the orientation contrast was reversed at a temporal frequency of 12.5 Hz, MS could nonetheless detect the contours defined by the orientation contrast despite being unable to judge whether the surfaces either side of the contrast were the same or not. This reveals that MS’s early visual cortex is sufficient for the intact phase-insensitive component of texture segmentation. Human area V4, therefore, is not necessary for texture segmentation.

Citation

Norman, L., Heywood, C., & Kentridge, R. (2017). Texture segmentation without human V4. Visual Cognition, 25(1-3), 184-195. https://doi.org/10.1080/13506285.2017.1301612

• Journal Article Type: Article
• Acceptance Date: Feb 21, 2017
• Online Publication Date: Mar 20, 2017
• Publication Date: Mar 20, 2017
• Deposit Date: Feb 28, 2017
• Publicly Available Date: Mar 20, 2018
• Journal: Visual Cognition
• Print ISSN: 1350-6285
• Electronic ISSN: 1464-0716
• Publisher: Taylor and Francis Group
• Peer Reviewed: Peer Reviewed
• Volume: 25
• Issue: 1-3
• Pages: 184-195
• DOI: https://doi.org/10.1080/13506285.2017.1301612
• Public URL: https://durham-repository.worktribe.com/output/1363423