Purchase this article with an account.
Barbara Dillenburger, Peter Kaskan, Haidong D. Lu, Anna W. Roe; Competition between real and illusory contours in early visual cortex revealed by a novel illusory contour stimulus (pure IC). Journal of Vision 2007;7(9):607. doi: 10.1167/7.9.607.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Visual scenes contain local (real) and global (illusory) contour cues which may signal conflicting and competing orientation signals. We previously suggested that these potential conflicts may be resolved by a competitive interaction between real and illusory contour processing in early visual cortex. By using subthreshold RCs superimposed on ICs, we demonstrated clear interference of IC perception by subthreshold RCs, thereby supporting a previous study suggesting suppression of RC signals in V1 by V2 feedback (Ramsden 2001). We now examine this hypothesis further using a novel IC stimulus.
We measured single-unit neural responses in anaesthetized macaque V1 and V2 to moving RCs embedded in four different contexts: RCalone, pureIC (Gu et al VSS 2006), KanizsaIC, and rotated pacmen (KanizsaControl). To examine orientation dependence, low contrast RCs were embedded parallel (superimposed upon) and orthogonal (abutting) to the IC. IC inducing cues were 2*RFwidths away from the RF.
We found opposite effects of pureIC and KanizsaIC contexts on RC response in V2. Whereas the pureIC context had a suppressive effect on parallel RC-responses in V2, the KanizsaIC resulted in facilitation of these V2 responses. Effects on parallel vs orthogonally oriented RCs also differed. Parallel RCs were suppressed (enhanced) by pureIC (KanizsaIC) contexts and enhanced (suppressed) by KanizsaIC (pureIC) contexts. In V1, effects were insignificant or reduced relative to V2 effects. Responses to the KanizsaControl were reduced or unchanged in V1 and V2.
Our results underscore a strong difference between illusory contour effects induced by pureICs vs KanizsaICs. We suggest that high contrast and collinearity cues present in Kanizsa figures confound the illusory contour contribution and result in misinterpretation of RC-IC interactions. In contrast to Dresp & Bonnet, our results are not consistent with IC-RC-summation. Importantly, our findings support roles for suppressive influences between parallel ICs and RCs, consistent with a competitive interaction.
This PDF is available to Subscribers Only