On perceived size

Abstract

It is a well-known fact that under certain circumstances one and the same object may be perceived as having different sizes. A common example of this phenomenon is the so-called moon illusion. In this paper a model is presented which is based on the Shannon theorem which appears to reproduce the data available for this type of observation very well. This theorem states that, for any data processing system, there is an upper-limit to the channel capacity. Assuming this to be valid for neurooptical systems, this means that there is a limited number of pixels available for image production; these pixels may be, however, distributed over a larger or smaller area of the retinal surface. Based on this assumption, a function is developed which determines the fraction of the total number of pixels available for the overall image which actually lie within the bounds of the object's image. This function thus gives a numerical value for the perceived size of the object of interest. In order to be able to apply this function it was calibrated by fitting the experimental data available from an early investigation of subjective perceived size. The result is another function which allows the prediction of the expected perceived size for situations similar to those which cause the real moon illusion and has been shown to fit two sets of experimental visual data.