Analysis of Visual Illusions Using Gabor Filters
Abstract
This thesis has demonstrated the correctness of using spatial filters in the analysis of various visual illusions; the specific form of filter has been derived from a Gabor equation model of simple cell response on the visual cortex. The Gabor Low Pass Filter (GLPF) applied to these anomalies was derived from proposals made that simple cell response on the visual cortex may be modeled by a set of equations originally proposed by Gabor in the 1940's. Based upon the extension of these equations into two dimensions, a GLPF process was applied to computer-generated black and white illusions (the Kanizsa Triangle, the Spoked Circle and the Ehrenstein Illusion). The results demonstrate that the anomalous contour present in these illusions are explained by an energy boundary surrounding the anomalous area. These differing energies are a direct result of the GLPF process. Much work has been accomplished attempting to understand how and why the human brain perceives visual illusions. From a physiological viewpoint, the process of vision may be divided into two main operations. The first of these deals with the processing which takes place when light first enters the eye up to the point when this information reaches the visual cortex of the brain. The second area regarding the concept of vision is concerned with how the brain uses this data-the cognition and conceptual formulation which allows us to recognize objects. This report looks at a viable way to model this processed data, via Gabor equations, and then turns toward possible explanations of certain categories of visual illusions.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1990
- Accession Number
- ADA230678
Entities
People
- Richard A. Oberndorf
Organizations
- Air Force Institute of Technology