Physiological Mechanisms for the Perception of Random Dot Moiré Patterns

Abstract

In 1968 I was working in the Department of Machine Intelligence and Perception at the University of Edinburgh with H.C. Longuet-Higgins. Longuet-Higgins returned from a trip to North America with a demonstration shown to him by E. Harth. Take a blank piece of paper and make a Xerox copy of it. Then take the copy and make a Xerox copy of it. If this process is iterated, then after not too long (by about the fifteenth generation using the 1968 Xerox machines) a dot pattern is generated which does not change significantly under further iterations [1]. I was interested in this demonstration, and realized that the process of random structure initiation plus long range inhibition could lead to stable spatial patterns [2–4]. In an effort to study the Xerox dot patterns, I made a transparency (to project on the wall using an overhead projector). The superposition of the transparency and the original led to the discovery of a Moiré effect. This effect is the subject of this note.