A new method for the restoration of two-dimensional (2-D) images obtained through a circularly band-limited system is given. Object and image are decomposed into circular harmonics, and it is observed that the imaging system acts separately on each harmonic. We show that superresolution is, in practice, attainable with a small number of one-dimensional iterations. The method presents several advantages on the conventional 2-D algorithms of the Gerchberg type. The computing effort in particular can be much reduced. Performances of our method on computer- generated images are presented.
© 1985 Optical Society of America
G. Gregori and S. Wabnitz, "New one-dimensional iterative approach to restoring two-dimensional circularly band-limited images," J. Opt. Soc. Am. A 2, 2046-2052 (1985)