I am working on a project where the purpose is to show the most common ocular ametropia (myopia, astigmatism and farsightedness) to lay people. One of the demonstrations I would like to do is: the user enters the data of his glasses (degrees and cylindrical and spherical axis) and an algorithm processes images from a database so that they are displayed as the individual sees without the glasses. Is there any mathematical modelling that can do this, without it being tough to implement? I read about "point spread function" using Fourier transform, but my teacher said it would do a masters job, and I do not have all that time. I thought of making an algorithm using lens association: measuring the diopter of a healthy human eye and associating lenses correcting the problem, but I do not know how to project the image that the lens would generate to simulate what the retina would receive from light .
1
-
1$\begingroup$ This application may require ray tracing. I suspect that the associated algorithm would be very difficult to develop in any short amount of time. Contact software developers who sell ray tracing programs and see if any of those programs have parameters in them that are adjustable. $\endgroup$– David WhiteCommented Apr 29, 2019 at 19:27
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
Researchers in this field (typically doctors studying the response of the human visual system) commonly use ray tracing programs to construct eye models. (Search for Gullstrand eye, Liou-Brennan eye model.). The eye model may have certain characteristics added that are dependent on one particular viewer (such as prescrption glass, or a uniquely shaped cornea, or an eyeball of a certain length. These unique characteristics may be found my measuring patients). You should realize that a huge part of the vision process is psychophysical - that is, you can determine what the image is on the retina, but what the brain "sees" is determined by neuorological processing. It's really hard to measure what the nerves and brain are doing.
