Abstract

Since the invention of superresolving pupils by Toraldo di Francia in 1952 [1] , superresolution has attracted the attention of many authors. As it is well known, the possibility of superresolution rests on some "a priori" information about the object. In the case of superresolving pupils, the only prior information is that the object has a finite extent. Elegant as they are from the theoretical point of view, the superresolving pupils [2] seem difficult to fabricate. As a matter of fact, rather severe tolerance conditions should be met in the fabrication and, to our knowledge, no practical superresolving pupil has been produced so far.

© 1986 Optical Society of America