期刊名称：Applied Computational Intelligence and Soft Computing
印刷版ISSN：1687-9724
电子版ISSN：1687-9732
出版年度：2017
卷号：2017
DOI：10.1155/2017/8179579
出版社：Hindawi Publishing Corporation
摘要：Geometrical illusions are a subclass of optical illusions in which the geometrical characteristics of patterns in particular orientations and angles are distorted and misperceived as a result of low-to-high-level retinal/cortical processing. Modelling the detection of tilt in these illusions, and its strength, is a challenging task and leads to the development of techniques that explain important features of human perception. We present here a predictive and quantitative approach for modelling foveal and peripheral vision for the induced tilt in the Café Wall illusion, in which parallel mortar lines between shifted rows of black and white tiles appear to converge and diverge. Difference of Gaussians is used to define a bioderived filtering model for the responses of retinal simple cells to the stimulus, while an analytical processing pipeline is developed to quantify the angle of tilt in the model and develop confidence intervals around them. Several sampling sizes and aspect ratios are explored to model variant foveal views, and a variety of pattern configurations are tested to model variant Gestalt views. The analysis of our model across this range of test configurations presents a precisely quantified comparison contrasting local tilt detection in the foveal sample sets with pattern-wide Gestalt tilt.
