A Computational Model of Visual Anisotropy

Bart Ons,Leopold Verstraelen,Johan Wagemans,Michael H Herzog

doi:10.1371/journal.pone.0021091

Abstract

Visual anisotropy has been demonstrated in multiple tasks where performance differs between vertical, horizontal, and oblique orientations of the stimuli. We explain some principles of visual anisotropy by anisotropic smoothing, which is based on a variation on Koenderink's approach in [1]. We tested the theory by presenting Gaussian elongated luminance profiles and measuring the perceived orientations by means of an adjustment task. Our framework is based on the smoothing of the image with elliptical Gaussian kernels and it correctly predicted an illusory orientation bias towards the vertical axis. We discuss the scope of the theory in the context of other anisotropies in perception.

Highlights

In a variety of tasks with stimuli presented at different orientations, like line orientation discrimination, e.g. [2], line vernier acuity and contrast sensitivity, e.g. [3,4], vertical and horizontal orientations have shown an advantage in performance compared to oblique orientations
From a visual inspection to the data, we found that the spread of responses was rather normally distributed but the magnitude of the variances differed for the different thicknesses of the Gaussian Luminance Profiles (GLPs)
A subtle difference in thickness of the luminance profile seemed to induce a strong effect on the magnitude of the perceived orientation bias

Summary

Introduction

In a variety of tasks with stimuli presented at different orientations, like line orientation discrimination, e.g. [2], line vernier acuity and contrast sensitivity, e.g. [3,4], vertical and horizontal orientations have shown an advantage in performance compared to oblique orientations. In the context of line orientation perception, it has been termed the ‘‘oblique effect’’, emphasizing the fact that the vertical and horizontal directions are discriminated better than the oblique ones [5,6,7,8,9,10,11]. We will first present the stimuli that seemed to induce an illusory orientation offset, we will demonstrate how this visual illusory perception of orientation can be related to anisotropy in resolution of the visual field. We will model the anisotropy in the resolution of the visual system by smoothing the initial image with elliptical Gaussian kernels instead, essentially, by smoothing more in the vertical direction than in the horizontal one. In the General Discussion section, we will elaborate on the proposed model on anisotropy and relate it to other issues in visual perception like symmetry detection, the ‘‘oblique’’ and the ‘‘horizontal’’ effect [see 12]

Methods

Results

Conclusion