Translator Disclaimer
Chapter 2:
Choice of Eye Models for Optical Evaluation
Chapter 1 assigned a clearly defined perspective to the optical analysis developed in this section. The goal is to fit the Campbell-Gubisch LSF data by means of a schematic eye model. The eye model for this task should be a finite one, characterized by aspherical surfaces and/or a graded index (GRIN) lens, to approximate the behavior of a real eye as closely as possible. The peculiar conditions of the Campbell-Gubisch experiment, namely double-pass foveal imaging of an axial object at a very low angular span (less than 1 deg), rule out any role played by odd aberrations (astigmatism, coma, etc.), thus reducing the need for an accurate reproduction of all monochromatic ocular aberrations over a wide range of conditions. For an integrated image element like LSF, attention is focused on spherical aberration, which can be reproduced by finite models with acceptable accuracy. Among the existing finite eye models cited in Chapter 1, none seem to have reached undisputed acceptance yet, in spite of a few significant differences existing in specific performances. Since a new eye model is unnecessary, two historical, exact schematic eyes are used, both due to Gullstrand. They are closely related to each other, because the only difference between them is in the lens modeling - a double-shell scheme in the first case and a graded-index scheme in the second, with all of the other parameters being equal. Both schematic eyes are transformed into finite models by providing them with suitable aspherical surfaces (as is generally done for deriving most finite models), and they are compared and contrasted in their capability of fitting the Campbell-Gubisch data.
Online access to SPIE eBooks is limited to subscribing institutions.

Back to Top