Purchase this article with an account.
Bruce C. Hansen, Robert F. Hess; Orientation tuning of contour integration. Journal of Vision 2007;7(9):605. doi: 10.1167/7.9.605.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
There currently exists an extensive body of literature devoted to understanding how the visual system integrates spatially segmented elements into contours, either artificially generated, or modeled after the contour statistics of natural scene imagery. However, little is known about the orientation tuning of the integration mechanism and whether such tuning changes as a function of contour curvature. To address those issues, we employed stimuli consisting of texture fields made up of pseudo-randomly distributed band-pass filtered noise elements (band-pass in spatial frequency and orientation), some of which, by virtue of their orientation alignment, formed a contour path. The orientation bandwidth of all filtered noise elements was varied, and the local spatial orientation misalignment (element-to-path angle) between the local element orientation and the contour path itself was systematically manipulated. The task consisted of a standard psychophysical 2AFC paradigm where observers were required to indicate which stimulus interval contained a contour. The results indicated that the local element orientation bandwidth needed to integrate low curvature contours was quite broad (ranging between 40° to 60°). For contours possessing a high degree of curvature, this bandwidth was significantly narrower (ranging between 20° to 30). However, the element-to-path angle varied very little as a function of contour curvature, ranging between 15° to 25° for all curvatures. The results indicate that the local element orientation tuning of the human visual contour integration mechanism is dependent on contour curvature.
This PDF is available to Subscribers Only